A method and device for generating a lake water surface area regulation scheme

Abstract

The application discloses a lake water surface area regulation scheme generation method and device. The application can accurately reflect the topography and water body properties of the research area by collecting lake basic data of the region to be regulated and constructing a high-precision digital elevation model (DEM) and a lake database; an initial simulation model is constructed by combining an environmental fluid dynamics model, water body boundary conditions, a DEM model and the lake database, which can fully consider the coupling influence of the water dynamic process and the topographic features on the lake water level and water surface distribution; after obtaining real-time monitoring data, the initial model is optimized based on the parameter prior distribution and the sample sampling strategy, which significantly improves the accuracy and response speed of the simulation model; a target simulation model is used to simulate various water level scenarios, response indexes under each scenario are obtained, and a regulation scheme is generated based on a multi-objective constraint curve and an analysis strategy, thereby improving the regulation accuracy of the lake water surface area.

Description

Technical Field

[0001] This invention relates to the field of lake regulation technology, and in particular to a method and apparatus for generating lake surface area regulation schemes. Background Technology

[0002] As vital water resource carriers and ecosystem units, the coordinated management of lake surface area and water quality is a core challenge for achieving sustainable water resource utilization. Traditional lake management often focuses on single-function needs such as flood control, water supply, and navigation, establishing empirical relationships between water level and area based on historical hydrological observation data to formulate water level control schemes based on statistical laws. However, with the intensification of global climate change and the deepening impact of human activities, problems such as lake eutrophication and ecological degradation are becoming increasingly prominent, and water level control models relying solely on hydrological indicators are no longer sufficient to meet the needs of water quality target management. In recent years, scholars at home and abroad have conducted extensive research in areas such as the development of hydrodynamic-water quality coupling models and ecological water level calculation, forming several typical technical approaches. However, these methods still have significant limitations in estimating suitable water surface area for water quality targets.

[0003] Existing methods for estimating lake surface area and regulating water level can be divided into three categories: First, regression models based on hydrological-topographic empirical formulas. Although computationally efficient, these models are essentially "black box" fittings based on historical observation data, failing to incorporate water quality evolution processes and making it difficult to reveal the coupling mechanism between changes in water surface area and pollutant transport. Second, static water quality-constrained water level optimization methods, while introducing local water quality indicators into the regulation model, typically only establish linear relationships for single pollutants and use fixed time-period thresholds, failing to construct continuous water level-multiple pollutant response surfaces and lacking adaptive capabilities to future pollution load fluctuations. Third, while separate hydrodynamic and water quality simulations achieve physical coupling, they often employ a two-step approach of "hydrodynamics first, then tracer," weakening the real-time interaction between biochemical processes such as pollutant degradation and sediment release and water flow. Furthermore, parameter calibration relies on manual trial and error, and sensitivity and uncertainty lack systematic quantification, making it prone to excellent local fitting but inaccurate global predictions under complex boundary conditions. Summary of the Invention

[0004] This invention provides a method and apparatus for generating lake water surface area control schemes to improve the accuracy of lake water surface area control.

[0005] To address the aforementioned technical problems, this invention provides a method for generating a lake surface area regulation scheme, comprising:

[0006] Collect basic lake data for the area to be regulated, and construct a DEM model and a lake database based on the basic lake data;

[0007] An initial lake simulation model is constructed based on the environmental fluid dynamics model, the water boundary conditions of the area to be regulated, the DEM model, and the lake database.

[0008] The real-time monitoring data of the area to be regulated is obtained, and the initial lake simulation model is optimized based on the preset parameter prior distribution, the preset sample sampling strategy and the real-time monitoring data to obtain the target lake simulation model.

[0009] Simulation is performed based on the target lake simulation model and preset water level scenario data to obtain response indicators corresponding to each water level scenario data. Based on the response indicators, preset multi-objective constraint curves, and preset analysis strategies, a lake surface area control scheme is generated.

[0010] This invention collects basic lake data from the area to be regulated and constructs a high-precision digital elevation model (DEM) and lake database, accurately reflecting the topography and water properties of the study area. An initial simulation model is constructed by combining environmental fluid dynamics models, water boundary conditions, the DEM model, and the lake database, fully considering the coupled influence of hydrodynamic processes and topographic features on lake water level and surface distribution. After acquiring real-time monitoring data, the initial model is optimized based on prior parameter distribution and sample sampling strategies, significantly improving the accuracy and response speed of the simulation model. The target simulation model is used to simulate various water level scenarios, obtaining response indicators under each scenario. Based on multi-objective constraint curves and analysis strategies, a control scheme is generated, achieving refined, multi-objective optimized control of the lake surface area, thereby improving the accuracy of lake surface area regulation.

[0011] Furthermore, the collection of basic lake data in the area to be regulated, and the construction of a DEM model and a lake database based on the basic lake data, includes:

[0012] Collect basic lake data for the area to be regulated, including vector boundary image data, underwater topographic point cloud data, hydrological and meteorological monitoring data, water quality spatiotemporal monitoring data, and pollutant inflow flux into the lake;

[0013] Outlier removal is performed on the time-series observation data, and the basic lake data, including vector boundary image data, underwater topographic point cloud data, hydrological and meteorological monitoring data, water quality spatiotemporal monitoring data, and pollutant inflow flux into the lake, are preprocessed to generate a spatial fusion database.

[0014] A DEM model is generated based on preset grid parameters, the vector boundary image data, and the underwater topographic point cloud data.

[0015] Based on a preset table structure and preset index specifications, the spatial fusion database and the DEM model are integrated to generate a lake database.

[0016] This invention constructs a high-quality spatial fusion database by simultaneously acquiring multi-source data, including vector boundary images, underwater topographic point clouds, hydrological and meteorological data, spatiotemporal water quality monitoring, and pollutant inflow flux into lakes. It also performs outlier removal and unified preprocessing on the time-series observation data, effectively eliminating the discreteness and noise interference of the original data. Based on preset grid parameters, the DEM model generated by fusing vector boundary images and underwater point cloud information accurately reflects the topographic undulations and water boundary characteristics of the lake, improving the accuracy and completeness of the lake topographic description. Finally, a unified table structure and indexing standard are used to seamlessly integrate the spatial fusion database and the DEM model, generating a structured and efficiently queryable lake database. This significantly improves data management efficiency and access performance, providing a solid and reliable data foundation for subsequent environmental fluid dynamics simulation, hydrological process analysis, and control decisions.

[0017] Furthermore, the construction of the initial lake simulation model based on the environmental fluid dynamics model, the water boundary conditions of the area to be regulated, the DEM model, and the lake database includes:

[0018] The region to be controlled is discretized in two planes based on the environmental fluid dynamics model and the curve orthogonal mesh method to construct a hydrodynamic model, and the vertical direction of the hydrodynamic model is layered based on the σ coordinate.

[0019] Adjust the water surface shear stress of each layer in the vertical direction of the hydrodynamic sub-model based on the hydrometeorological monitoring data;

[0020] The bottom friction coefficients of each layer in the vertical direction of the hydrodynamic sub-model are set based on the underwater topographic point cloud data.

[0021] The Coriolis force coefficients of the hydrodynamic sub-model are set based on the geographical latitude of the region to be regulated;

[0022] Based on the aforementioned spatiotemporal water quality monitoring data, COD sub-model, TN sub-model, TP sub-model, and DO sub-model are constructed, and a water quality model is constructed based on the aforementioned COD sub-model, TN sub-model, TP sub-model, and DO sub-model;

[0023] An initial lake simulation model is constructed based on the water body boundary conditions, the water quality model, and the hydrodynamic model.

[0024] This invention employs a curved orthogonal grid method to discretize the area to be regulated in two-dimensional planes and layers it based on σ coordinates. This allows for precise characterization of the flow characteristics at different depths of the lake while ensuring numerical stability. It dynamically adjusts the shear stress at each water surface layer using hydrological and meteorological monitoring data and accurately sets the bottom friction coefficient based on underwater topographic point clouds, enhancing the responsiveness of the hydrodynamic sub-model to actual boundary conditions and bed characteristics. The invention incorporates geographic latitude information to set the Coriolis force coefficient, achieving accurate simulation of large-scale rotational effects. Based on spatiotemporal water quality monitoring data, COD, TN, TP, and DO sub-models are constructed and integrated into the water quality model, enabling the initial lake simulation model to simultaneously reflect hydrodynamic processes and capture the laws of material transport and transformation. Finally, by integrating water boundary conditions, the hydrodynamic model, and the water quality model, a hydrodynamic and water quality coupled initial lake simulation model is formed.

[0025] Furthermore, the preset prior parameter distributions include the COD decay coefficient distribution, TP sedimentation rate distribution, and nitrification rate coefficient distribution; the preset sample sampling strategy is a Latin hypercube sampling strategy; the acquisition of real-time monitoring data of the area to be regulated, and the optimization of the initial lake simulation model based on the preset prior parameter distributions, the preset sample sampling strategy, and the real-time monitoring data to obtain the target lake simulation model, includes:

[0026] Several sets of parameter combinations are generated based on the Latin hypercube sampling strategy and the COD decay coefficient distribution, TP sedimentation rate distribution and nitrification rate coefficient distribution.

[0027] The real-time monitoring data of the area to be regulated is obtained, the real-time monitoring parameters are input into the initial lake simulation model, and the initial lake simulation model is run based on each set of parameter combinations to obtain the accuracy index corresponding to the set of parameter combinations. The optimal parameter set is obtained based on the accuracy index.

[0028] Based on the optimal parameter set and the preset kernel density estimation method, the posterior distribution of the parameters is determined, and the initial lake simulation model is optimized based on the posterior distribution of the parameters to obtain the target lake simulation model.

[0029] This invention introduces the prior distributions of three key water quality parameters—COD decay coefficient, TP sedimentation rate, and nitrification rate—and employs a Latin hypercube sampling strategy to generate multiple sets of high-dimensional parameter combinations. This enables uniform coverage of the model input space with a limited number of samples, significantly improving parameter optimization efficiency. Combined with real-time monitoring data-driven model operation and accuracy evaluation, it ensures that the selection of the optimal parameter set fully reflects the on-site hydrodynamic and water quality evolution characteristics. Furthermore, the kernel density estimation method is used to extract the posterior distribution of parameters, effectively quantifying the uncertainty of model parameters. Based on this, the initial simulation model is optimized, which not only improves the computational accuracy and robustness of the target lake simulation model but also enhances its rapid response capability to sudden environmental disturbances, providing reliable model support for the accurate formulation of subsequent water surface area control schemes.

[0030] Furthermore, the water level scenario data includes several water level values; it includes the simulation based on the target lake simulation model and the preset water level scenario data to obtain the response index corresponding to each water level scenario data, and determining the optimal water level value and optimal water surface area based on the response index, the preset multi-objective constraint curve, and the preset analysis strategy, including:

[0031] The target lake simulation model is run based on each water level value until the target lake simulation model reaches a steady state. The response indicators corresponding to each water level value are obtained. The response indicators include the distribution of pollutant concentrations in the whole lake and the proportion of areas that meet the standards.

[0032] Based on the water level scenario data and the distribution of pollutant concentrations across the entire lake and the proportion of areas meeting standards corresponding to each water level value, a water level compliance area response relationship is constructed.

[0033] Based on the water level compliance area response relationship, the multi-objective constraint curve, and the preset analysis strategy, a feasible water level set is determined.

[0034] The prediction accuracy of the water level values ​​in the feasible water level set is calculated based on flood season conditions and non-flood season conditions respectively. A lake water surface area regulation scheme is generated based on the prediction accuracy. The lake water surface area regulation scheme includes the optimal water level value, the optimal water surface area, the regulation time window, and the pollution control threshold.

[0035] This invention obtains key response indicators such as the distribution of pollutant concentrations across the entire lake and the proportion of areas meeting standards by running target simulation models at each preset water level value until a steady state is reached. Based on this, a water level-standard-area response relationship is constructed, enabling precise quantification of the impact of water level changes on lake pollution control. Combining multi-objective constraint curves and preset analysis strategies, feasible water level sets that balance pollution control and water surface area requirements are selected, ensuring the feasibility of the generated control scheme under multiple objectives. Predictive accuracy assessments of the feasible water level sets under both flood and non-flood season conditions fully reflect the impact of different hydrological scenarios on control effects, improving the robustness of the strategy. Finally, a lake water surface area control scheme including optimal water level value, optimal water surface area, control time window, and pollution control threshold is output. This not only provides a quantitative decision-making basis for pollutant concentration compliance and ecological protection but also provides clear and operable parameters and implementation sequences for actual water resource allocation and environmental management.

[0036] Secondly, the present invention provides a device for generating a lake surface area regulation scheme, comprising: a data acquisition module, a model building model, a simulation module, and a scheme generation module;

[0037] The data acquisition module is used to collect basic lake data in the area to be regulated, and to construct a DEM model and a lake database based on the basic lake data;

[0038] The model construction model is used to build an initial lake simulation model based on the environmental fluid dynamics model, the water boundary conditions of the area to be regulated, the DEM model, and the lake database.

[0039] The simulation module is used to acquire real-time monitoring data of the area to be regulated, and optimize the initial lake simulation model based on the preset parameter prior distribution, the preset sample sampling strategy and the real-time monitoring data to obtain the target lake simulation model.

[0040] The scheme generation module is used to perform simulation based on the target lake simulation model and preset water level scenario data, obtain response indicators corresponding to each water level scenario data, and generate a lake water surface area control scheme based on the response indicators, preset multi-objective constraint curves and preset analysis strategies.

[0041] Furthermore, the data acquisition module is used to collect basic lake data of the area to be regulated, and to construct a DEM model and a lake database based on the basic lake data, including:

[0042] Collect basic lake data for the area to be regulated, including vector boundary image data, underwater topographic point cloud data, hydrological and meteorological monitoring data, water quality spatiotemporal monitoring data, and pollutant inflow flux into the lake;

[0043] Outlier removal is performed on the time-series observation data, and the basic lake data, including vector boundary image data, underwater topographic point cloud data, hydrological and meteorological monitoring data, water quality spatiotemporal monitoring data, and pollutant inflow flux into the lake, are preprocessed to generate a spatial fusion database.

[0044] A DEM model is generated based on preset grid parameters, the vector boundary image data, and the underwater topographic point cloud data.

[0045] Based on a preset table structure and preset index specifications, the spatial fusion database and the DEM model are integrated to generate a lake database.

[0046] Furthermore, the model construction model is used to construct an initial lake simulation model based on the environmental fluid dynamics model, the water boundary conditions of the area to be regulated, the DEM model, and the lake database, including:

[0047] The region to be controlled is discretized in two planes based on the environmental fluid dynamics model and the curve orthogonal mesh method to construct a hydrodynamic model, and the vertical direction of the hydrodynamic model is layered based on the σ coordinate.

[0048] Adjust the water surface shear stress of each layer in the vertical direction of the hydrodynamic sub-model based on the hydrometeorological monitoring data;

[0049] The bottom friction coefficients of each layer in the vertical direction of the hydrodynamic sub-model are set based on the underwater topographic point cloud data.

[0050] The Coriolis force coefficients of the hydrodynamic sub-model are set based on the geographical latitude of the region to be regulated;

[0051] Based on the aforementioned spatiotemporal water quality monitoring data, COD sub-model, TN sub-model, TP sub-model, and DO sub-model are constructed, and a water quality model is constructed based on the aforementioned COD sub-model, TN sub-model, TP sub-model, and DO sub-model;

[0052] An initial lake simulation model is constructed based on the water body boundary conditions, the water quality model, and the hydrodynamic model.

[0053] Furthermore, the preset prior parameter distributions include COD decay coefficient distribution, TP sedimentation rate distribution, and nitrification rate coefficient distribution; the preset sample sampling strategy is a Latin hypercube sampling strategy; the simulation module is used to acquire real-time monitoring data of the area to be regulated, and optimize the initial lake simulation model based on the preset prior parameter distributions, the preset sample sampling strategy, and the real-time monitoring data to obtain the target lake simulation model, including:

[0054] Several sets of parameter combinations are generated based on the Latin hypercube sampling strategy and the COD decay coefficient distribution, TP sedimentation rate distribution and nitrification rate coefficient distribution.

[0055] The real-time monitoring data of the area to be regulated is obtained, the real-time monitoring parameters are input into the initial lake simulation model, and the initial lake simulation model is run based on each set of parameter combinations to obtain the accuracy index corresponding to the set of parameter combinations. The optimal parameter set is obtained based on the accuracy index.

[0056] Based on the optimal parameter set and the preset kernel density estimation method, the posterior distribution of the parameters is determined, and the initial lake simulation model is optimized based on the posterior distribution of the parameters to obtain the target lake simulation model.

[0057] Furthermore, the scheme generation module is used for water level scenario data including several water level values; including performing simulation based on the target lake simulation model and preset water level scenario data, obtaining response indicators corresponding to each water level scenario data, and determining the optimal water level value and optimal water surface area based on the response indicators, preset multi-objective constraint curves, and preset analysis strategies, including:

[0058] The target lake simulation model is run based on each water level value until the target lake simulation model reaches a steady state. The response indicators corresponding to each water level value are obtained. The response indicators include the distribution of pollutant concentrations in the whole lake and the proportion of areas that meet the standards.

[0059] Based on the water level scenario data and the distribution of pollutant concentrations across the entire lake and the proportion of areas meeting standards corresponding to each water level value, a water level compliance area response relationship is constructed.

[0060] Based on the water level compliance area response relationship, the multi-objective constraint curve, and the preset analysis strategy, a feasible water level set is determined.

[0061] The prediction accuracy of the water level values ​​in the feasible water level set is calculated based on flood season conditions and non-flood season conditions respectively. A lake water surface area regulation scheme is generated based on the prediction accuracy. The lake water surface area regulation scheme includes the optimal water level value, the optimal water surface area, the regulation time window, and the pollution control threshold. Attached Figure Description

[0062] Figure 1 This is a schematic flowchart of a method for generating a lake water surface area regulation scheme according to an embodiment of the present invention;

[0063] Figure 2 This is a schematic diagram of a device for generating a lake surface area regulation scheme, provided in an embodiment of the present invention. Detailed Implementation

[0064] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and are not intended to limit the scope of the invention.

[0065] The terms "first" and "second," etc., in the specification, claims, and drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such processes, methods, products, or apparatus.

[0066] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0067] Example 1

[0068] See Figure 1 , Figure 1 This is a schematic flowchart illustrating a method for generating a lake surface area control scheme according to an embodiment of the present invention. The embodiment of the present invention provides a method for generating a lake surface area control scheme, including steps 101 to 104, as detailed below:

[0069] Step 101: Collect basic lake data for the area to be regulated, and construct a DEM model and lake database based on the basic lake data;

[0070] In this embodiment, the process of collecting basic lake data for the area to be regulated, and constructing a DEM model and a lake database based on the basic lake data, includes:

[0071] Collect basic lake data for the area to be regulated, including vector boundary image data, underwater topographic point cloud data, hydrological and meteorological monitoring data, water quality spatiotemporal monitoring data, and pollutant inflow flux into the lake;

[0072] Outlier removal is performed on the time-series observation data, and the basic lake data, including vector boundary image data, underwater topographic point cloud data, hydrological and meteorological monitoring data, water quality spatiotemporal monitoring data, and pollutant inflow flux into the lake, are preprocessed to generate a spatial fusion database.

[0073] A DEM model is generated based on preset grid parameters, the vector boundary image data, and the underwater topographic point cloud data.

[0074] Based on a preset table structure and preset index specifications, the spatial fusion database and the DEM model are integrated to generate a lake database.

[0075] In this embodiment, basic lake data for the area to be regulated is collected. This basic lake data is multi-dimensional and heterogeneous, including: synchronously acquiring high-resolution satellite imagery and deriving the lake shoreline vector boundary; collecting underwater topographic point clouds using LiDAR or echo sounders; integrating hydrological and meteorological monitoring data such as water level, wind speed and direction, and inflow river flow; covering spatiotemporal monitoring data of water quality such as COD, TP, TN, and DO from conventional cross-sections and automatic stations; and establishing a daily pollutant inflow flux database based on a distributed SCS hydrological model and pollution source survey.

[0076] In this embodiment, outlier removal is performed on all monitoring sequences in the lake basic data (based on statistical thresholds or box plot methods); missing data is filled using spatial kriging interpolation and time series interpolation algorithms; and a spatial fusion database that has been consistent and integrates data from all sources is generated.

[0077] In this embodiment, a regular grid is generated by calling an interpolation algorithm (TIN or IDW) based on preset grid parameters (such as 10–50 m grid spacing), shoreline vector and point cloud height values; in areas without actual measurement points, the elevation is completed by ArcGIS spatial interpolation and 3D reconstruction tools based on historical waterway maps and monitoring section data; thereby generating a digital elevation model (DEM) that meets the accuracy requirements.

[0078] In this embodiment, a unified table structure and indexing standard are designed, covering spatial elements (shoreline, DEM) and attribute elements (hydrology, water quality, flux); the spatial fusion database and DEM data are imported into a relational or spatiotemporal database, and spatial and attribute indexes are established; a lake database that can be efficiently retrieved, visualized, and exported is output.

[0079] In this embodiment, the focus is on constructing a model input database and employing multi-source heterogeneous data fusion technology to achieve unified management of spatial and attribute data. First, the lake shoreline vector boundary is obtained through high-resolution satellite image interpretation. Combined with underwater topographic data collected by LiDAR or echo sounders, a digital elevation model (DEM) with a spatial resolution of 10-50 meters is constructed. For areas lacking measured topographic data, ArcGIS spatial interpolation technology is used to reconstruct historical navigation maps and hydrological station cross-sectional data in three dimensions. Regarding hydrological data acquisition, water level monitoring data from the water resources department, wind speed and direction records from meteorological stations, and inflow river flow monitoring data are integrated to establish a continuous dataset with a time step of 1 hour. Water quality data acquisition covers conventional monitoring sections and automatic monitoring stations, focusing on acquiring spatiotemporal distribution data of core indicators such as chemical oxygen demand (COD), total phosphorus (TP), total nitrogen (TN), and dissolved oxygen (DO). Simultaneously, sediment sampling is conducted to analyze heavy metal content and pollutant release rates. Pollution load calculation employs a combination of a distributed SCS hydrological model and pollution source survey data. Inputs such as point source emissions, non-point source runoff, and atmospheric deposition are classified and quantified to construct a daily pollutant inflow flux database. The data preprocessing stage utilizes outlier removal and missing data imputation algorithms to statistically correct outliers in the monitoring data and fill in data gaps in the monitoring areas using spatial kriging interpolation.

[0080] This invention constructs a high-quality spatial fusion database by simultaneously acquiring multi-source data, including vector boundary images, underwater topographic point clouds, hydrological and meteorological data, spatiotemporal water quality monitoring, and pollutant inflow flux into lakes. It also performs outlier removal and unified preprocessing on the time-series observation data, effectively eliminating the discreteness and noise interference of the original data. Based on preset grid parameters, the DEM model generated by fusing vector boundary images and underwater point cloud information accurately reflects the topographic undulations and water boundary characteristics of the lake, improving the accuracy and completeness of the lake topographic description. Finally, a unified table structure and indexing standard are used to seamlessly integrate the spatial fusion database and the DEM model, generating a structured and efficiently queryable lake database. This significantly improves data management efficiency and access performance, providing a solid and reliable data foundation for subsequent environmental fluid dynamics simulation, hydrological process analysis, and control decisions.

[0081] Step 102: Construct an initial lake simulation model based on the environmental fluid dynamics model, the water boundary conditions of the area to be regulated, the DEM model, and the lake database;

[0082] In this embodiment, the construction of the initial lake simulation model based on the environmental fluid dynamics model, the water boundary conditions of the area to be regulated, the DEM model, and the lake database includes:

[0083] The region to be controlled is discretized in two planes based on the environmental fluid dynamics model and the curve orthogonal mesh method to construct a hydrodynamic model, and the vertical direction of the hydrodynamic model is layered based on the σ coordinate.

[0084] Adjust the water surface shear stress of each layer in the vertical direction of the hydrodynamic sub-model based on the hydrometeorological monitoring data;

[0085] The bottom friction coefficients of each layer in the vertical direction of the hydrodynamic sub-model are set based on the underwater topographic point cloud data.

[0086] The Coriolis force coefficients of the hydrodynamic sub-model are set based on the geographical latitude of the region to be regulated;

[0087] Based on the aforementioned spatiotemporal water quality monitoring data, COD sub-model, TN sub-model, TP sub-model, and DO sub-model are constructed, and a water quality model is constructed based on the aforementioned COD sub-model, TN sub-model, TP sub-model, and DO sub-model;

[0088] An initial lake simulation model is constructed based on the water body boundary conditions, the water quality model, and the hydrodynamic model.

[0089] In this embodiment, high-precision two-dimensional curved orthogonal grid discretization and vertical σ-coordinate layering, combined with multi-source monitoring data and measured boundary conditions, achieve a tight coupling between hydrodynamic and water quality elements. Simultaneously, by fully utilizing real-time meteorological, underwater topographic, and water quality monitoring information, key parameters such as surface shear stress, bottom friction coefficient, and Coriolis force are dynamically set to ensure that the model can accurately simulate the lake flow field and pollutant migration process under different water levels and pollution loads.

[0090] In this embodiment, the two-dimensional planar domain of the lake is discretized based on the curvilinear orthogonal horizontal grid supported by the EFDC model to adapt to complex lake shore morphology and improve boundary fitting accuracy; in the vertical direction, σ coordinates are used to layer (usually 3-5 layers) to realize the analysis of laminar flow characteristics at different depths of shallow lakes.

[0091] In this embodiment, the grid size is dynamically adjusted according to the lake morphology, and key areas (such as sewage outlets and wetland boundaries) are locally densified to a resolution of 20 meters.

[0092] In this embodiment, the wind stress term uses wind speed data at a height of 10 meters, which is converted into water surface shear stress using a vertical wind speed correction formula; the bottom friction coefficient is assigned values ​​according to the bottom type (silt, sand, rock), with a range of 0.02-0.035 m. 1 / 3 / s; The Coriolis force coefficient is dynamically calculated based on the geographical latitude of the lake.

[0093] In this embodiment, four molecular models for COD, TN, TP, and DO were constructed. COD degradation adopted a first-order kinetic equation with a temperature correction factor of 1.047 to reflect the temperature dependence of organic matter oxidation rate. A sediment release module was introduced to couple the nitrogen and phosphorus cycle process. Its release rate parameters were determined by laboratory culture and field monitoring inversion to enhance the model's ability to reproduce sediment source pollution.

[0094] In this embodiment, the rivers flowing into the lake are input using measured flow rates and pollutant concentration time series. Discharge outlets are generalized as point source emissions. Non-point source pollution is calculated using a distributed SCS model to determine runoff and correlate it with land use type pollution load coefficients. Initial model conditions are set using water quality survey data from the dry season. For areas lacking historical data, measured river flow rates, point source pollution, and non-point source pollution load series are imported, using the dry season water quality survey as initial values. A spin-up simulation is then performed until the system reaches steady state, outputting the final initial conditions.

[0095] This invention employs a curved orthogonal grid method to discretize the area to be regulated in two-dimensional planes and layers it based on σ coordinates. This allows for precise characterization of the flow characteristics at different depths of the lake while ensuring numerical stability. It dynamically adjusts the shear stress at each water surface layer using hydrological and meteorological monitoring data and accurately sets the bottom friction coefficient based on underwater topographic point clouds, enhancing the responsiveness of the hydrodynamic sub-model to actual boundary conditions and bed characteristics. The invention incorporates geographic latitude information to set the Coriolis force coefficient, achieving accurate simulation of large-scale rotational effects. Based on spatiotemporal water quality monitoring data, COD, TN, TP, and DO sub-models are constructed and integrated into the water quality model, enabling the initial lake simulation model to simultaneously reflect hydrodynamic processes and capture the laws of material transport and transformation. Finally, by integrating water boundary conditions, the hydrodynamic model, and the water quality model, a hydrodynamic and water quality coupled initial lake simulation model is formed.

[0096] Step 103: Obtain real-time monitoring data of the area to be regulated, optimize the initial lake simulation model based on the preset parameter prior distribution, the preset sample sampling strategy and the real-time monitoring data, and obtain the target lake simulation model;

[0097] In this embodiment, the preset prior parameter distribution includes the COD decay coefficient distribution, TP sedimentation rate distribution, and nitrification rate coefficient distribution; the preset sample sampling strategy is a Latin hypercube sampling strategy; the step of acquiring real-time monitoring data of the area to be regulated, and optimizing the initial lake simulation model based on the preset prior parameter distribution, the preset sample sampling strategy, and the real-time monitoring data to obtain the target lake simulation model includes:

[0098] Several sets of parameter combinations are generated based on the Latin hypercube sampling strategy and the COD decay coefficient distribution, TP sedimentation rate distribution and nitrification rate coefficient distribution.

[0099] The real-time monitoring data of the area to be regulated is obtained, the real-time monitoring parameters are input into the initial lake simulation model, and the initial lake simulation model is run based on each set of parameter combinations to obtain the accuracy index corresponding to the set of parameter combinations. The optimal parameter set is obtained based on the accuracy index.

[0100] Based on the optimal parameter set and the preset kernel density estimation method, the posterior distribution of the parameters is determined, and the initial lake simulation model is optimized based on the posterior distribution of the parameters to obtain the target lake simulation model.

[0101] In this embodiment, firstly, based on literature review and localized experiments, the prior distribution range of key parameters was determined. The chemical oxygen demand (COD) decay coefficient was set to 0.05-0.15 day⁻¹, the total phosphorus (TP) sedimentation rate to 0.01-0.03 m / day, and the nitrification rate coefficient to 0.1-0.3 day⁻¹, covering the typical characteristics of lake organic matter degradation, particulate matter sedimentation, and nitrogen cycle processes. Subsequently, Latin hypercube sampling (LHS) was used to generate 500-1000 uniformly distributed parameter combinations in the three-dimensional parameter space to avoid sampling bias inherent in traditional Monte Carlo methods.

[0102] In this embodiment, during model optimization, real-time monitoring data (including indicators such as water level, COD, TP, and dissolved oxygen, with a time resolution ≤ 1 hour) are input into the initial lake simulation model. Simulation scenarios with all parameter combinations are run in parallel, and the accuracy indices for each parameter group are calculated: the Nash-Sutcliffe Efficiency (NSE) is used to evaluate the overall goodness of fit between simulated and measured values ​​(target threshold NSE > 0.6), and the Root Mean Square Error (RMSE) quantifies the range of absolute deviation. By screening a set of high-likelihood parameters that meet the NSE threshold, a Behavioral Uncertainty Set (BUS) is constructed, and the Kernel Density Estimation (KDE) method is used to analyze the posterior distribution of the parameters. Finally, the median is selected as the calibration value to reduce the impact of extreme values ​​on model stability.

[0103] In this embodiment, to verify the spatial adaptability of parameter optimization, a moving window method is introduced to divide the lake into 1km × 1km verification units, and the correlation coefficient between simulated and measured values ​​is calculated for each unit. For areas with errors exceeding 30%, the spatial consistency of the model is improved by correcting local grid parameters (such as adjusting the bottom friction coefficient or settlement rate).

[0104] This invention introduces the prior distributions of three key water quality parameters—COD decay coefficient, TP sedimentation rate, and nitrification rate—and employs a Latin hypercube sampling strategy to generate multiple sets of high-dimensional parameter combinations. This enables uniform coverage of the model input space with a limited number of samples, significantly improving parameter optimization efficiency. Combined with real-time monitoring data-driven model operation and accuracy evaluation, it ensures that the selection of the optimal parameter set fully reflects the on-site hydrodynamic and water quality evolution characteristics. Furthermore, the kernel density estimation method is used to extract the posterior distribution of parameters, effectively quantifying the uncertainty of model parameters. Based on this, the initial simulation model is optimized, which not only improves the computational accuracy and robustness of the target lake simulation model but also enhances its rapid response capability to sudden environmental disturbances, providing reliable model support for the accurate formulation of subsequent water surface area control schemes.

[0105] Step 104: Based on the target lake simulation model and preset water level scenario data, perform simulation to obtain the response indicators corresponding to each water level scenario data, and generate a lake water surface area control scheme based on the response indicators, preset multi-objective constraint curves and preset analysis strategies.

[0106] In this embodiment, the water level scenario data includes several water level values; the process includes simulating based on the target lake simulation model and preset water level scenario data to obtain response indicators corresponding to each water level scenario data, and determining the optimal water level value and optimal water surface area based on the response indicators, preset multi-objective constraint curves, and preset analysis strategies, including:

[0107] The target lake simulation model is run based on each water level value until the target lake simulation model reaches a steady state. The response indicators corresponding to each water level value are obtained. The response indicators include the distribution of pollutant concentrations in the whole lake and the proportion of areas that meet the standards.

[0108] Based on the water level scenario data and the distribution of pollutant concentrations across the entire lake and the proportion of areas meeting standards corresponding to each water level value, a water level compliance area response relationship is constructed.

[0109] Based on the water level compliance area response relationship, the multi-objective constraint curve, and the preset analysis strategy, a feasible water level set is determined.

[0110] The prediction accuracy of the water level values ​​in the feasible water level set is calculated based on flood season conditions and non-flood season conditions respectively. A lake water surface area regulation scheme is generated based on the prediction accuracy. The lake water surface area regulation scheme includes the optimal water level value, the optimal water surface area, the regulation time window, and the pollution control threshold.

[0111] In this embodiment, a water level regulation scenario matrix is ​​set up based on verifying the accuracy of the model. Using historical water level fluctuations as a benchmark, water level sequences are generated at 0.2-meter intervals (e.g., 15 scenarios from dead water level to flood control limit water level). Each scenario maintains a constant inflow of pollution load into the lake, and the model is run until steady state (usually requiring 30-60 hydrological days).

[0112] In this embodiment, the average concentrations of COD, TP, and TN of the entire lake and the proportion of areas meeting the standards (i.e., the percentage of areas meeting Class III water quality standards) corresponding to each water level value are obtained, and a water level-compliant area response relationship is established through cubic polynomial regression.

[0113] (1)

[0114] Where A is the area that meets the standard, H is the water level, and a, b, c, and d are regression coefficients.

[0115] In this embodiment, a water level-compliant area response equation is constructed using cubic polynomial regression to quantify the dynamic impact of water level changes on the area of ​​water quality compliant. Its regression coefficient ( a , b , c , d The goodness of fit was determined by the least squares method. R 2) The value needs to be higher than 0.85 to ensure the reliability of the model.

[0116] In this embodiment, a two-dimensional response surface methodology (RSM) multi-objective optimization problem is constructed simultaneously. For example, when the combined constraints of COD ≤ 20 mg / L and TP ≤ 0.05 mg / L are simultaneously met, the Pareto front analysis method is used to determine the optimal water level range. For lakes with multiple discharge outlets, a pollution source intensity control variable is added. Thousands of pollution reduction schemes and water level combinations are generated through Monte Carlo simulation, and a random forest algorithm is used to screen feasible solutions that meet the water quality objectives.

[0117] In this embodiment, to address the need for coordinated control of multiple pollutants, Pareto front analysis is employed to screen for non-dominated solutions within a two-dimensional response surface (e.g., a composite constraint of COD ≤ 20 mg / L and TP ≤ 0.05 mg / L) to determine the optimal water level range. For lakes with multiple discharge outlets, a pollution source intensity control variable is added. Thousands of pollution reduction schemes and water level combinations are generated through Monte Carlo simulation, and a random forest algorithm is used to rank the importance of features (e.g., node splitting contribution > 5%) to screen for feasible solutions that meet the water quality objectives.

[0118] In this embodiment, seasonal differences are incorporated into the regulation strategy: the flood season model focuses on the dilution effect of high water levels on pollutants, while the non-flood season model is coupled with a sediment release kinetics module (the release rate is calibrated through indoor incubation experiments) to independently calculate the optimal water level value.

[0119] In this embodiment, the actual water surface area at typical water levels is extracted based on historical remote sensing images, and an absolute error analysis (required to be <5%) is performed between the actual water surface area and the model-estimated value. Next, temporary monitoring sections are set up, and water quality is continuously monitored for 30 days at the selected control water level to verify the accuracy of the predicted proportion of compliant area (allowable error ±10%). The final comprehensive control scheme includes the optimal water surface area, the control time window (e.g., high water level dilution during the flood season from July to September), and the pollution control threshold (e.g., TN ≤ 1.5 mg / L). This method, through water level-water quality dynamic response modeling, multi-objective optimization algorithms, and seasonal adaptability verification, significantly improves the engineering applicability and ecological sustainability of lake water surface area control schemes, providing a quantitative decision-making tool for lake management under complex hydrological conditions.

[0120] This invention obtains key response indicators such as the distribution of pollutant concentrations across the entire lake and the proportion of areas meeting standards by running target simulation models at each preset water level value until a steady state is reached. Based on this, a water level-standard-area response relationship is constructed, enabling precise quantification of the impact of water level changes on lake pollution control. Combining multi-objective constraint curves and preset analysis strategies, feasible water level sets that balance pollution control and water surface area requirements are selected, ensuring the feasibility of the generated control scheme under multiple objectives. Predictive accuracy assessments of the feasible water level sets under both flood and non-flood season conditions fully reflect the impact of different hydrological scenarios on control effects, improving the robustness of the strategy. Finally, a lake water surface area control scheme including optimal water level value, optimal water surface area, control time window, and pollution control threshold is output. This not only provides a quantitative decision-making basis for pollutant concentration compliance and ecological protection but also provides clear and operable parameters and implementation sequences for actual water resource allocation and environmental management.

[0121] This invention collects basic lake data from the area to be regulated and constructs a high-precision digital elevation model (DEM) and lake database, accurately reflecting the topography and water properties of the study area. An initial simulation model is constructed by combining environmental fluid dynamics models, water boundary conditions, the DEM model, and the lake database, fully considering the coupled influence of hydrodynamic processes and topographic features on lake water level and surface distribution. After acquiring real-time monitoring data, the initial model is optimized based on prior parameter distribution and sample sampling strategies, significantly improving the accuracy and response speed of the simulation model. The target simulation model is used to simulate various water level scenarios, obtaining response indicators under each scenario. Based on multi-objective constraint curves and analysis strategies, a control scheme is generated, achieving refined, multi-objective optimized control of the lake surface area, thereby improving the accuracy of lake surface area regulation.

[0122] In this embodiment, a dynamic response mechanism between water quality targets and water surface area is established to achieve multi-objective collaborative optimization of lake management. Compared with existing technologies, it has two significant advantages: first, the GLUE-BMA joint algorithm is used to improve the model's adaptability under complex boundary conditions; second, a two-dimensional response surface is constructed to solve the nonlinear optimization problem of collaborative control of multiple pollutants.

[0123] In this embodiment, by constructing a coupled model of the entire process involving hydrodynamics, biochemistry, and sediment interaction, the limitations of simulating a single physical process are overcome. Taking sediment pollutant release as an example, existing technologies often use fixed rate parameters, while this invention innovatively introduces a dynamic release model: through indoor culture experiments, the synergistic effects of temperature, dissolved oxygen, and hydraulic shear stress on the sediment phosphorus release rate are quantified, and a comprehensive release equation including an exponential temperature correction term, an oxygen-dependent inhibition function, and a shear stress factor is established (as shown in equation: ), where R base The model uses the base release rate, T as water temperature, DO as dissolved oxygen concentration, and τ as bed shear stress. This mechanism-driven modeling approach allows water surface area control to not only reflect the physical dilution effect of water volume but also capture the intrinsic contribution of biochemical processes to water quality improvement.

[0124] In this embodiment, traditional single-objective optimization methods struggle to handle the conflicts and synergies among multiple pollutant indicators. This invention, however, achieves global optimal solution search under multiple constraints by constructing a high-dimensional response surface and using Pareto front analysis. Specifically, for indicators such as COD, TP, TN, and DO, tens of thousands of water level-pollution load combinations are generated using Latin hypercube sampling. Parallel computation is then used to obtain multi-objective outputs such as the proportion of the lake's compliant area, peak pollutant concentrations, and ecological replenishment costs. An adaptive grid refinement algorithm rapidly converges to the Pareto optimal solution set within the objective space, and a random forest model is used to rank the importance of each control variable. Furthermore, a chance-constrained programming model is introduced, allowing decision-makers to set compliance probability thresholds (e.g., requiring ≥80% TP compliance area at 95% confidence level), and the system automatically selects a set of solutions that meet their risk preferences. This multi-objective optimization capability upgrades water surface area control from "single-indicator optimization" to intelligent decision-making based on "multi-benefit equilibrium."

[0125] In this embodiment, traditional models treat lakes as homogeneous water bodies, leading to water quality deterioration in high-risk areas such as areas near sewage outlets and lake bay retention areas being masked by the overall compliance rate. This invention achieves a detailed characterization of microscale processes through high-resolution grid division (locally refined to 20 meters) and spatial heterogeneity parameter assignment. During the model construction phase, sub-meter-level lakebed topographic data is acquired using UAV aerial surveying and side-scan sonar to identify sediment pollution hotspots and set release parameters individually. The validation phase introduces a moving window analysis method, dividing the lake into 1km × 1km validation units, and independently calculating the NSE and relative error index for each unit. For areas with errors exceeding 30% (such as the central lake area strongly influenced by wind-driven currents), a local parameter optimization module based on Bayesian inversion is activated, adjusting only parameters such as the bottom friction coefficient and pollutant diffusion rate in that area, rather than global recalibration. In addition, an algorithm for assessing the spatial confidence of compliant areas was developed. Combined with Kriging interpolation to generate probability distribution maps, the dynamic changes of high-risk areas under different water levels can be displayed intuitively, providing targeted guidance for the precise implementation of projects such as ecological dredging and artificial wetland construction.

[0126] In this embodiment, addressing the shortcomings of traditional parameter calibration, such as strong subjectivity and unclear error propagation, this invention constructs a GLUE-BMA joint uncertainty analysis framework. First, 1000 parameter combinations (covering key parameters such as COD decay coefficient, TP settling rate, and nitrification intensity) are generated through Latin hypercube sampling. The EFDC model is then run in parallel to calculate the probability distribution of compliant area under each scenario. Simultaneously, a sensitivity contribution analysis module is developed, using the Sobol index method to quantify the influence weight of each input parameter on the compliant area. For example, the global sensitivity index of TP settling rate is found to be 0.78, guiding on-site monitoring to prioritize strengthening sediment characteristic sampling. This shift from "point estimation" to "probability distribution" allows decision-makers to clearly understand the risk boundaries of the scheme and avoid misjudgments caused by overconfidence in the model.

[0127] Please refer to Figure 2 , Figure 2 A schematic diagram of a lake water surface area regulation scheme generation device provided in an embodiment of the present invention includes: a data acquisition module 201, a model construction module 202, a simulation module 203, and a scheme generation module 204;

[0128] The data acquisition module 201 is used to collect basic lake data of the area to be regulated, and to construct a DEM model and a lake database based on the basic lake data.

[0129] The model construction model 202 is used to construct an initial lake simulation model based on the environmental fluid dynamics model, the water boundary conditions of the area to be regulated, the DEM model, and the lake database.

[0130] The simulation module 203 is used to acquire real-time monitoring data of the area to be regulated, and optimize the initial lake simulation model based on the preset parameter prior distribution, the preset sample sampling strategy and the real-time monitoring data to obtain the target lake simulation model.

[0131] The scheme generation module 204 is used to perform simulation based on the target lake simulation model and preset water level scenario data, obtain response indicators corresponding to each water level scenario data, and generate a lake water surface area control scheme based on the response indicators, preset multi-objective constraint curves and preset analysis strategies.

[0132] Furthermore, the data acquisition module is used to collect basic lake data of the area to be regulated, and to construct a DEM model and a lake database based on the basic lake data, including:

[0133] Collect basic lake data for the area to be regulated, including vector boundary image data, underwater topographic point cloud data, hydrological and meteorological monitoring data, water quality spatiotemporal monitoring data, and pollutant inflow flux into the lake;

[0134] Outlier removal is performed on the time-series observation data, and the basic lake data, including vector boundary image data, underwater topographic point cloud data, hydrological and meteorological monitoring data, water quality spatiotemporal monitoring data, and pollutant inflow flux into the lake, are preprocessed to generate a spatial fusion database.

[0135] A DEM model is generated based on preset grid parameters, the vector boundary image data, and the underwater topographic point cloud data.

[0136] Based on a preset table structure and preset index specifications, the spatial fusion database and the DEM model are integrated to generate a lake database.

[0137] Furthermore, the model construction model is used to construct an initial lake simulation model based on the environmental fluid dynamics model, the water boundary conditions of the area to be regulated, the DEM model, and the lake database, including:

[0138] The region to be controlled is discretized in two planes based on the environmental fluid dynamics model and the curve orthogonal mesh method to construct a hydrodynamic model, and the vertical direction of the hydrodynamic model is layered based on the σ coordinate.

[0139] Adjust the water surface shear stress of each layer in the vertical direction of the hydrodynamic sub-model based on the hydrometeorological monitoring data;

[0140] The bottom friction coefficients of each layer in the vertical direction of the hydrodynamic sub-model are set based on the underwater topographic point cloud data.

[0141] The Coriolis force coefficients of the hydrodynamic sub-model are set based on the geographical latitude of the region to be regulated;

[0142] Based on the aforementioned spatiotemporal water quality monitoring data, COD sub-model, TN sub-model, TP sub-model, and DO sub-model are constructed, and a water quality model is constructed based on the aforementioned COD sub-model, TN sub-model, TP sub-model, and DO sub-model;

[0143] An initial lake simulation model is constructed based on the water body boundary conditions, the water quality model, and the hydrodynamic model.

[0144] Furthermore, the preset prior parameter distributions include COD decay coefficient distribution, TP sedimentation rate distribution, and nitrification rate coefficient distribution; the preset sample sampling strategy is a Latin hypercube sampling strategy; the simulation module is used to acquire real-time monitoring data of the area to be regulated, and optimize the initial lake simulation model based on the preset prior parameter distributions, the preset sample sampling strategy, and the real-time monitoring data to obtain the target lake simulation model, including:

[0145] Several sets of parameter combinations are generated based on the Latin hypercube sampling strategy and the COD decay coefficient distribution, TP sedimentation rate distribution and nitrification rate coefficient distribution.

[0146] The real-time monitoring data of the area to be regulated is obtained, the real-time monitoring parameters are input into the initial lake simulation model, and the initial lake simulation model is run based on each set of parameter combinations to obtain the accuracy index corresponding to the set of parameter combinations. The optimal parameter set is obtained based on the accuracy index.

[0147] Based on the optimal parameter set and the preset kernel density estimation method, the posterior distribution of the parameters is determined, and the initial lake simulation model is optimized based on the posterior distribution of the parameters to obtain the target lake simulation model.

[0148] Furthermore, the scheme generation module is used for water level scenario data including several water level values; including performing simulation based on the target lake simulation model and preset water level scenario data, obtaining response indicators corresponding to each water level scenario data, and determining the optimal water level value and optimal water surface area based on the response indicators, preset multi-objective constraint curves, and preset analysis strategies, including:

[0149] The target lake simulation model is run based on each water level value until the target lake simulation model reaches a steady state. The response indicators corresponding to each water level value are obtained. The response indicators include the distribution of pollutant concentrations in the whole lake and the proportion of areas that meet the standards.

[0150] Based on the water level scenario data and the distribution of pollutant concentrations across the entire lake and the proportion of areas meeting standards corresponding to each water level value, a water level compliance area response relationship is constructed.

[0151] Based on the water level compliance area response relationship, the multi-objective constraint curve, and the preset analysis strategy, a feasible water level set is determined.

[0152] The prediction accuracy of the water level values ​​in the feasible water level set is calculated based on flood season conditions and non-flood season conditions respectively. A lake water surface area regulation scheme is generated based on the prediction accuracy. The lake water surface area regulation scheme includes the optimal water level value, the optimal water surface area, the regulation time window, and the pollution control threshold.

[0153] Example 2

[0154] Daihai Lake, an important semi-arid lake in northern my country, has faced severe ecological problems in recent years, including a continuous decline in water level and aggravated eutrophication. By applying a method for generating lake surface area control schemes, the ecological water replenishment strategy was optimized and the water environment quality was improved. The specific implementation steps and effect verification are as follows:

[0155] In this embodiment, a comprehensive monitoring network for Daihai Lake was first constructed, integrating satellite remote sensing, ground observation, and experimental analysis data. For topographic data, an airborne LiDAR was used to acquire a digital elevation model (DEM) of the lake basin with a spatial resolution of 1 meter, identifying sediment-rich areas (mainly distributed in the northwestern bay area) and hard substrate areas (southern shallows). Hydrological and meteorological data were integrated from hourly wind speed and rainfall records from five surrounding national meteorological stations. Water temperature, dissolved oxygen, COD, TP, and TN parameters were collected in real time using buoy-type water quality monitoring stations (located in the lake center, downstream of the northern sewage outlet, and at the southern wetland entrance), with a sampling frequency of 15 minutes per station. Pollution load calculation employed a distributed SCS model, combined with land use data (30-meter resolution) and an agricultural non-point source pollution coefficient manual, to calculate the annual total phosphorus load entering Daihai Lake as 86.3 tons, with agricultural runoff contributing 62%. Sediment characteristics were analyzed using columnar sampling (20 points in total) and indoor incubation experiments to determine the phosphorus release rate of sediment at different temperatures (5-30℃) and dissolved oxygen levels, and a dynamic release equation was established.

[0156] (2)

[0157] Where R P denoted as phosphorus release rate (mg / m²·d), T as water temperature (°C), DO as dissolved oxygen at the bottom layer (mg / L), and τ as bed surface shear stress (Pa).

[0158] In this embodiment, based on the EFDC model framework, a curved orthogonal grid is used to spatially discretize Daihai Lake. The basic grid size is 50 meters, which is refined to 20 meters in the northwestern sedimentary area, with four vertical σ-coordinate layers. In the hydrodynamic module, the wind stress coefficient is set to 1.2 × 10⁻⁶. −6 The bottom friction coefficient is assigned a value based on the bottom sediment type: 0.025 m for silt zones. 1 / 3 / s, sandy area 0.032m 1 / 3 / s. The water quality module couples the dynamic sediment release equation with the phytoplankton uptake model, where the algal growth rate parameter is calibrated using MODIS chlorophyll a product (spatial resolution 500 meters), and the maximum growth rate is set at 1.2 d. −1 The boundary conditions include four main rivers flowing into the lake (flow range 0.5-8.2 m³ / s) and three sewage outlets (daily average COD load 12.6 tons). Non-point source pollution is input through the runoff-concentration relationship curve generated by the SCS model.

[0159] In this embodiment, the GLUE method was used for model parameter calibration, and monitoring data from 36 months of data collected between 2018 and 2020 were selected for training. 1200 parameter combinations were generated using Latin hypercube sampling. The prior distribution of the key parameter was: COD decay coefficient (0.05-0.15 d...-1 ), TP settling rate (0.01-0.03 m / d), nitrification rate (0.1-0.4 d). −1 A set of 248 qualified parameters with a Nash efficiency coefficient (NSE) > 0.6 was selected. Posterior analysis showed that TP settling rate had the most significant impact on the compliant area (Sobol index 0.71). The median of the final calibration parameters was: COD decrease of 0.08 days. −1 TP settling rate: 0.018 m / d; nitrification rate: 0.25 d. −1 The validation phase used an independent dataset from 2021, and the NSE of the simulated COD and TP values ​​for the entire lake reached 0.82 and 0.76, respectively, which is more than 35% higher than the traditional fixed parameter model.

[0160] In this embodiment, a water level regulation scenario matrix is ​​set up, with the historical lowest water level of Daihai Lake (1203.5 meters, 2019) and the highest water level (1206.2 meters, 2005) as boundaries, generating 15 water level scenarios at 0.2-meter intervals. Each scenario fixes the inflow pollution load to the measured average value in 2021, and the model is run to steady state (approximately 30 hydrological days). The output results extract the average COD and TP concentrations and the proportion of areas meeting standards (Class III water standard: COD ≤ 20 mg / L, TP ≤ 0.05 mg / L) for the entire lake and its sub-regions (northern industrial impact area, central open area, and southern wetland buffer zone). A water level-standard-meeting area response equation is established through cubic polynomial regression.

[0161] A COD =-215.4H 3 +1296.8H 2 -2587.5H+1723.1 (R) 2 =0.94) (3)

[0162] A TP =184.2H 3 -1102.6H 2 +2198.3H-1456.7 (R) 2 =0.89) (4)

[0163] Where A represents the compliant area (km²) and H represents the water level (meters). The analysis shows that when the water level rises from 1204.0 meters to 1205.0 meters, the COD compliant area increases from 58.7% to 76.3%, while the TP compliant area decreases from 63.4% to 51.2%, reflecting a significant trade-off among multiple objectives.

[0164] In this embodiment, the NSGA-II algorithm is used to solve for the Pareto optimal solution set, with the objective function set as maximizing the combined COD and TP compliance rate and minimizing the ecological water replenishment volume. Through 100,000 iterations, 120 sets of non-dominated solutions are generated. The optimal compromise is a water level of 1204.8 meters, corresponding to a water replenishment volume of 120 million m³ / year, achieving a COD compliance rate of 74.2% and a TP compliance rate of 68.5%. Compared to the traditional single-objective solution (water level 1205.2 meters, water replenishment volume 180 million m³), ​​this achieves 33% water saving while maintaining a combined compliance rate reduction of only 3%. Furthermore, a cost function is introduced (water replenishment cost 0.35 yuan / m³, wastewater treatment cost 2.8 yuan / kg TP). Random forest feature importance analysis reveals that agricultural non-point source emission reduction contributes 57%, guiding the priority implementation of lakeside buffer wetland projects.

[0165] In this embodiment, comparative experiments and data verification were conducted to quantify the advantages of the present invention. Two sets of comparative experiments were set up: ① Traditional empirical formula method: The cubic polynomial model of Xiong Jiaqing (2016) was used to calculate the water surface area with the same water level data, without considering water quality constraints. The results showed that when the water level was 1204.8 meters, the estimated area deviated from that of the present invention by only 2.1%, but the corresponding TP compliance area was overestimated by 19.6% (because it assumes that the area expansion will necessarily improve water quality). ② Static parameter model: The EFDC model with a fixed bottom sediment release rate (0.015 mg / m²·d) was used. In the low-temperature season (water temperature 8℃), the simulated TP concentration was underestimated by up to 37%, while the error of the lake water surface area regulation scheme generation method described in this embodiment was controlled within 9%.

[0166] In this embodiment of the invention, a terminal device is also provided, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the above-described method for generating a lake surface area control scheme.

[0167] In this embodiment of the invention, a computer-readable storage medium is also provided, which includes a stored computer program, wherein the computer program controls the device where the computer-readable storage medium is located to execute the above-described method for generating a lake surface area regulation scheme when it is running.

[0168] For example, a computer program can be divided into one or more modules, one or more of which are stored in memory and executed by a processor to perform the present invention. The one or more modules can be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in a terminal device.

[0169] The terminal device may be a desktop computer, laptop, handheld computer, or cloud server, etc. The terminal device may include, but is not limited to, a processor, memory, and display. Those skilled in the art will understand that the above components are merely examples of terminal devices and do not constitute a limitation on the terminal device. It may include more or fewer components, or combinations of certain components, or different components. For example, the terminal device may also include input / output devices, network access devices, buses, etc.

[0170] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the terminal device, connecting all parts of the terminal device through various interfaces and lines.

[0171] Memory can be used to store computer programs and / or modules. The processor implements various functions of the terminal device by running or executing the computer programs and / or modules stored in the memory, and by accessing data stored in the memory. Memory can mainly include a program storage area and a data storage area. The program storage area can store the operating system, application programs required for at least one function (such as sound playback, text conversion, etc.), etc.; the data storage area can store data created based on the use of the mobile phone (such as audio data, text message data, etc.). In addition, memory can include high-speed random access memory, and can also include non-volatile memory, such as hard disks, RAM, plug-in hard disks, smart media cards (SMC), secure digital cards (SD cards), flash cards, at least one disk storage device, flash memory device, or other volatile solid-state storage devices.

[0172] The module generated based on the lake surface area control scheme, if implemented as a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. Those skilled in the art can understand and implement this without any inventive effort.

[0173] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A method for generating a lake water surface area regulation scheme, characterized in that, The method comprises the following steps: collecting lake basic data of a region to be controlled, and constructing a DEM model and a lake database based on the lake basic data; constructing an initial lake simulation model based on an environmental fluid dynamics model, water body boundary conditions of the region to be controlled, the DEM model and the lake database; obtaining real-time monitoring data of the region to be controlled, and optimizing the initial lake simulation model based on a preset parameter prior distribution, a preset sample sampling strategy and the real-time monitoring data to obtain a target lake simulation model; based on the target lake simulation model and preset water level scenario data, performing simulation and emulation to obtain response indexes corresponding to each water level scenario data, and generating a lake water surface area control scheme based on the response indexes, a preset multi-objective constraint curve and a preset analysis strategy.

2. The method of claim 1, wherein, The method comprises the following steps: collecting lake basic data of a region to be controlled, and constructing a DEM model and a lake database based on the lake basic data; collecting lake basic data of a region to be controlled, wherein the lake basic data comprises vector boundary image data, underwater topographic point cloud data, hydro-meteorological monitoring data, water quality spatiotemporal monitoring data and pollutant inflow flux into a lake; performing outlier rejection processing on the lake basic data, and pre-processing the lake basic data to generate a spatial fusion database; generating a DEM model based on preset grid parameters, the vector boundary image data and the underwater topographic point cloud data; 3. The method of claim 2, wherein, integrating the spatial fusion database and the DEM model based on a preset table structure and a preset index specification to generate a lake database. The method comprises the following steps: performing planar two-dimensional discretization on the region to be controlled based on an environmental fluid dynamics model and a curvilinear orthogonal grid method to construct a hydrodynamic model, and layering the hydrodynamic model in a vertical direction based on a sigma coordinate; adjusting water surface shear stress of each layer in the vertical direction of the hydrodynamic model based on the hydro-meteorological monitoring data; setting bottom friction coefficients of each layer in the vertical direction of the hydrodynamic model based on the underwater topographic point cloud data; setting Coriolis force coefficients of the hydrodynamic model based on geographical latitude of the region to be controlled; constructing COD, TN, TP and DO sub-models based on the water quality spatiotemporal monitoring data, and constructing a water quality model based on the COD, TN, TP and DO sub-models; 4. The method of claim 3, wherein, constructing an initial lake simulation model based on the water body boundary conditions, the water quality model and the hydrodynamic model. The preset parameter prior distribution comprises COD decay coefficient distribution, TP sedimentation rate distribution and nitrification rate coefficient distribution; the preset sample sampling strategy is a Latin hypercube sampling strategy; and the method comprises the following steps: generating a plurality of sets of parameter combinations based on the Latin hypercube sampling strategy and the COD decay coefficient distribution, TP settling rate distribution, and nitrification rate coefficient distribution; obtaining real-time monitoring data of the to-be-controlled region, inputting the real-time monitoring parameters into the initial lake simulation model, and running the initial lake simulation model based on each set of parameter combinations respectively to obtain precision indicators corresponding to each set of parameter combinations, and obtaining an optimal parameter set based on the precision indicators; determining a parameter posterior distribution based on the optimal parameter set and a preset kernel density estimation method, optimizing the initial lake simulation model based on the parameter posterior distribution, and obtaining a target lake simulation model.

5. The method of claim 4, wherein, The water level scenario data includes a plurality of water level values; based on the target lake simulation model and the preset water level scenario data, simulation is performed to obtain response indicators corresponding to each water level scenario data, and a lake water surface area control scheme is generated based on the response indicators, a preset multi-objective constraint curve, and a preset analysis strategy, including: running the target lake simulation model based on each water level value until the target lake simulation model reaches a steady state, obtaining response indicators corresponding to each water level value, the response indicators including lake-wide pollutant concentration distribution and compliance area proportion; constructing a water level compliance area response relationship based on the water level scenario data and the lake-wide pollutant concentration distribution and compliance area proportion corresponding to each water level value; determining a feasible water level set based on the water level compliance area response relationship, the multi-objective constraint curve, and the preset analysis strategy; calculating prediction accuracy for water level values in the feasible water level set based on flood period conditions and non-flood period conditions, and generating a lake water surface area control scheme based on the prediction accuracy, the lake water surface area control scheme including an optimal water level value, an optimal water surface area, a control time window, and a pollution control threshold.

6. A lake water surface area regulation scheme generation device characterized by comprising: including: a data acquisition module, a model construction module, a simulation module, and a scheme generation module; The data acquisition module is configured to acquire lake basic data of a to-be-controlled region, and construct a DEM model and a lake database based on the lake basic data; The model construction module is configured to construct an initial lake simulation model based on an environmental fluid dynamics model, water boundary conditions of the to-be-controlled region, the DEM model, and the lake database; The simulation module is configured to obtain real-time monitoring data of the to-be-controlled region, optimize the initial lake simulation model based on a preset parameter prior distribution, a preset sample sampling strategy, and the real-time monitoring data, and obtain a target lake simulation model; The scheme generation module is configured to perform simulation based on the target lake simulation model and preset water level scenario data to obtain response indicators corresponding to each water level scenario data, and generate a lake water surface area control scheme based on the response indicators, a preset multi-objective constraint curve, and a preset analysis strategy.

7. A lake water surface area regulation scheme generation device according to claim 6, wherein, The data acquisition module is configured to acquire lake basic data of a to-be-controlled region, and construct a DEM model and a lake database based on the lake basic data, including: Collecting lake basic data of a region to be regulated, the lake basic data comprising vector boundary image data, underwater topography point cloud data, hydrological meteorological monitoring data, water quality spatiotemporal monitoring data and pollutant inflow flux into a lake; Performing outlier rejection processing on the lake basic data and preprocessing the lake basic data to generate a spatial fusion database; Generating a DEM model based on preset grid parameters, the vector boundary image data and the underwater topography point cloud data; Integrating the spatial fusion database and the DEM model based on a preset table structure and a preset index specification to generate a lake database.

8. A lake water surface area regulation scheme generation device according to claim 7, wherein, The model construction model is configured to construct an initial lake simulation model based on an environmental fluid dynamics model, water body boundary conditions of the region to be regulated, a DEM model and a lake database, and comprises: Performing planar two-dimensional discretization on the region to be regulated based on the environmental fluid dynamics model and a curved orthogonal grid method, constructing a hydrodynamic model, and layering the vertical direction of the hydrodynamic model based on a sigma coordinate; Adjusting the water surface shear stress of each layer in the vertical direction of the hydrodynamic model based on the hydrological meteorological monitoring data; Setting the bottom friction coefficient of each layer in the vertical direction of the hydrodynamic model based on the underwater topography point cloud data; Setting the Coriolis force coefficient of the hydrodynamic model based on the geographical latitude of the region to be regulated; Constructing COD, TN, TP and DO submodels based on the water quality spatiotemporal monitoring data, and constructing a water quality model based on the COD, TN, TP and DO submodels; Constructing an initial lake simulation model based on the water body boundary conditions, the water quality model and the hydrodynamic model.

9. A lake water surface area regulation scheme generation device according to claim 8, wherein, The preset parameter prior distribution comprises a COD decay coefficient distribution, a TP settling rate distribution and a nitrification rate coefficient distribution; the preset sample sampling strategy is a Latin hypercube sampling strategy; the simulation module is configured to acquire real-time monitoring data of the region to be regulated, optimize the initial lake simulation model based on a preset parameter prior distribution, a preset sample sampling strategy and the real-time monitoring data, and acquire a target lake simulation model, and comprises: Generating a plurality of groups of parameter combinations based on the Latin hypercube sampling strategy and the COD decay coefficient distribution, the TP settling rate distribution and the nitrification rate coefficient distribution; Acquiring real-time monitoring data of the region to be regulated, inputting the real-time monitoring parameters into the initial lake simulation model, running the initial lake simulation model based on each group of parameter combinations respectively, acquiring precision indexes corresponding to each group of parameter combinations, and acquiring an optimal parameter set based on the precision indexes; Determining a parameter posterior distribution based on the optimal parameter set and a preset kernel density estimation method, optimizing the initial lake simulation model based on the parameter posterior distribution, and acquiring a target lake simulation model.

10. A lake water surface area regulation scheme generation device according to claim 9, wherein, The water level scenario data comprises a plurality of water level values; the scheme generation module is configured to simulate based on the target lake simulation model and preset water level scenario data, obtain a response index corresponding to each water level scenario data, and generate a lake water surface area regulation scheme based on the response index, a preset multi-objective constraint curve and a preset analysis strategy, comprising: running the target lake simulation model based on each water level value until the target lake simulation model reaches a steady state, obtaining a response index corresponding to each water level value, the response index comprising a lake-wide pollutant concentration distribution and a compliance area ratio; constructing a water level compliance area response relationship based on the water level scenario data and the lake-wide pollutant concentration distribution and the compliance area ratio corresponding to each water level value; determining a feasible water level set based on the water level compliance area response relationship, the multi-objective constraint curve and the preset analysis strategy; calculating a prediction accuracy based on the water level values in the feasible water level set under flood period conditions and non-flood period conditions, and generating a lake water surface area regulation scheme based on the prediction accuracy, the lake water surface area regulation scheme comprising an optimal water level value, an optimal water surface area, a regulation time window and a pollution control threshold.

Images