Lake flooding prevention and disposal method based on water power regulation

Abstract

This invention discloses a method for the prevention and control of lacustrine floods based on hydrodynamic regulation, comprising: constructing a comprehensive risk assessment index for lacustrine floods by acquiring multi-source environmental data to define the spatial risk gradient distribution of high-incidence areas; using a three-dimensional hydrodynamic-aquatic ecology coupled model, with dissolved oxygen concentration safety threshold as a constraint, to reverse-engineer target hydrodynamic regulation parameters; and determining the deployment density and spatial location of the propulsion devices accordingly. During the operation phase, the instantaneous dynamic risk status is determined by combining real-time monitoring data, and the real-time operating power of the propulsion devices is continuously adjusted accordingly. This invention achieves full life-cycle management from static risk assessment and physical demand deduction to dynamic servo regulation, enabling precise targeted prevention and control of lacustrine flood outbreak risks, effectively preventing the formation of anaerobic environments while reducing energy consumption.

Description

Technical Field

[0001] This invention relates to the field of lake water environment management technology, and in particular to a method for preventing and treating lake flooding based on hydrodynamic regulation. Background Technology

[0002] Lacustrine flooding (also known as black water mass) is a severe ecological disaster in eutrophic lakes caused by the combined effects of cyanobacterial blooms and accumulation, sediment release, and anaerobic environments under specific meteorological and hydrological conditions. Preventing lacustrine flooding is of significant technical importance for ensuring the ecological security of shallow lakes and maintaining water quality stability. By disrupting the static stratification of the water body through hydrodynamic regulation and enhancing the reoxygenation capacity of the bottom layer, the proliferation of anaerobic microorganisms can be inhibited mechanistically, and the sulfur-iron reaction process can be blocked. This is the key physical control method for achieving precise prevention and emergency response to lacustrine flooding.

[0003] Currently, the prevention and control of lake floods typically employs methods such as pollution source control and interception, ecological dredging, or chemical aeration. In terms of physical regulation, existing technologies mainly improve water flow by deploying propulsion or aeration equipment in predetermined areas. These devices are usually deployed based on experience, with point coverage, and their operation often employs fixed power or simple control logic based on single-factor thresholds. However, the causes of lake floods in large lakes exhibit significant spatial heterogeneity and temporal fluctuations. Traditional regulation methods often struggle to accurately adapt to the complex patterns of water risk evolution, leading to bottlenecks in practical operational applications such as uneven resource allocation, excessive energy consumption, and unstable control effects.

[0004] Therefore, the common challenge facing existing technologies lies in overcoming the shortcomings of insufficient targeting and low efficiency in hydrodynamic regulation, and solving the problems of blindness in spatial layout and lag in time response. There is an urgent need for a method that can effectively coordinate static layout planning and dynamic execution strategies to improve the scientific rigor and adaptability of lake flood control in complex environmental conditions. Summary of the Invention

[0005] The purpose of this invention is to provide a method for the prevention and treatment of lacustrine flooding based on hydrodynamic regulation, in order to solve the aforementioned problems in the prior art.

[0006] Technical solution: A method for preventing and managing lacustrine flooding based on hydrodynamic regulation, comprising:

[0007] Acquire multi-source environmental data for the target lake;

[0008] Based on multi-source environmental data, a comprehensive risk assessment index for lacustrine flooding of the target lake is constructed, and the spatial risk gradient distribution of high-incidence lacustrine flooding areas in the target lake is defined accordingly.

[0009] Using a three-dimensional hydrodynamic-aquatic ecology coupled model of the target lake, and with a preset safe threshold for dissolved oxygen concentration as a constraint, the target hydrodynamic control parameters required to bring the dissolved oxygen concentration to the safe threshold in the high-incidence area of ​​lake flooding are reverse-engineered.

[0010] Based on the spatial risk gradient distribution and target hydrodynamic control parameters, the deployment density of propulsion devices in high-incidence lacustrine flooding areas and the spatial deployment locations of each propulsion device are determined.

[0011] Real-time monitoring data of areas prone to lacustrine flooding is obtained, and the instantaneous dynamic risk status of these areas is determined based on the real-time monitoring data. The real-time operating power of each propulsion device is then adjusted accordingly.

[0012] Optional, multi-source environmental data include water quality monitoring data, hydro-meteorological data, and sediment monitoring data;

[0013] Before constructing the comprehensive risk assessment index for lacustrine biota, the following data preprocessing steps are also included:

[0014] Based on multi-source environmental data, we extract the basic risk factors of the target lake in each spatial grid cell.

[0015] The basic risk factors are normalized and uniformly mapped to dimensionless risk scores.

[0016] Optionally, a comprehensive risk assessment index for algal blooms of the target lake may be constructed, including:

[0017] For any spatial grid cell, based on its normalized basic risk factors, calculate the main effect risk contribution when each individual risk factor acts independently.

[0018] Obtain pre-screened risk factor pairs with significant positive synergistic amplification effect, and calculate the synergistic interaction risk contribution of the risk factor pairs based on the product of the scores when the two basic risk factors co-occur.

[0019] Based on preset interaction intensity adjustment parameters, the risk contribution of the main effect and the risk contribution of the synergistic interaction are nonlinearly weighted and fused to obtain the comprehensive risk assessment index for lake pansmography.

[0020] Optionally, after obtaining the comprehensive risk assessment index for lacustrine flooding, and before defining the spatial risk gradient distribution of high-incidence areas in the target lake, a correction for flow-induced risk propagation is also included:

[0021] Based on the three-dimensional hydrodynamic-aquatic ecology coupled model of the target lake, the baseline net flux of water flow between adjacent spatial grid cells in the target lake is extracted, and the upstream connected neighborhood of each spatial grid cell is determined accordingly.

[0022] Based on the comprehensive risk assessment index of lake flooding in each spatial grid unit, and according to the preset connectivity influence coefficient, the risk increment of each unit in the upstream connected neighborhood after flux ratio weighting is iteratively accumulated to the local spatial grid unit until the iteration converges, and the final comprehensive risk assessment index of lake flooding after spatial propagation correction is obtained.

[0023] Spatial clustering was performed based on the final comprehensive risk assessment index for lacustrine flooding to determine the continuous boundaries of high-incidence areas of lacustrine flooding and their spatial risk gradient distribution.

[0024] Optionally, the instantaneous dynamic risk status of areas prone to lacustrine flooding can be determined based on real-time monitoring data, including:

[0025] Obtain the preset absolute safety threshold and preset extreme danger threshold corresponding to each monitoring indicator in the real-time monitoring data;

[0026] Using segmented risk conversion logic, the real-time physical measurement values ​​of each monitoring indicator are mapped to dimensionless instantaneous risk scores. Measurement values ​​that are at or above the preset absolute safety threshold are mapped to zero risk scores, while measurement values ​​that are at or below the preset extreme danger threshold are mapped to full risk scores.

[0027] Obtain the dynamic calculation weights of each pre-configured monitoring indicator;

[0028] The instantaneous risk scores of each monitoring indicator are weighted and summed with their corresponding dynamic calculation weights to obtain the instantaneous dynamic risk status that characterizes the sensitivity to multi-factor comprehensive triggering.

[0029] Optionally, adjust the real-time operating power of each propulsion device, including:

[0030] Obtain pre-configured low-risk response thresholds and high-risk warning thresholds;

[0031] Determine the position of the instantaneous dynamic risk status within different warning intervals divided by low-risk response thresholds and high-risk warning thresholds;

[0032] Calculate the dynamic adjustment ratio factor for smooth transition by using continuous adjustment logic that matches the landing warning interval;

[0033] The reference operating power of each propulsion device is obtained, and the reference operating power is multiplied by the dynamic adjustment ratio factor to set the real-time operating power of each propulsion device.

[0034] Optionally, multi-source environmental data also includes remote sensing monitoring data; extracting various basic risk factors for the target lake in each spatial grid cell, including:

[0035] From multi-source environmental data, the frequency factors of low oxygen in water, high temperature, stable weather, endogenous risk factors of sediment containing organic matter and sulfur and iron content, and high-density aggregation frequency factors of cyanobacteria detected by remote sensing images were extracted as basic risk factors.

[0036] Optionally, the propulsion devices are arranged in a layered manner at their spatial locations, including:

[0037] Surface propulsion devices are installed on the surface of the water body at the target spatial locations to improve the horizontal flow of the water; and,

[0038] At the target spatial location, a bottom-level propulsion device is installed in the bottom layer of the water body to break up the vertical static stratification of the water body and enhance the reoxygenation capacity of the bottom water body.

[0039] Optionally, the propulsion device is also equipped with an aeration device based on the Venturi principle;

[0040] Adjusting the real-time operating power of each propulsion device also includes:

[0041] While controlling the operation of the propulsion device, the aeration equipment is driven to draw in air using the negative pressure generated by the high-speed water flow, forming microbubble dissolved air water, which, together with the hydrodynamic circulation of the propulsion device, improves the overall reoxygenation efficiency of the water body in the high-incidence area of ​​lacustrine flooding.

[0042] A system for preventing and managing lacustrine flooding based on hydrodynamic regulation, comprising:

[0043] At least one processor; and,

[0044] A memory that is communicatively connected to at least one processor; wherein,

[0045] The memory stores instructions that can be executed by a processor to implement the steps of any one of the above-described methods for preventing and controlling lacustrine flooding based on hydrodynamic regulation.

[0046] A computer-readable storage medium comprising a stored executable program.

[0047] Specifically, when the executable program is running, it controls the device containing the computer-readable storage medium to perform any of the steps described in the above-mentioned method for preventing and controlling lacustrine flooding based on hydrodynamic regulation.

[0048] Beneficial effects: This invention realizes full life cycle management from static risk assessment and physical demand deduction to dynamic servo control, and can carry out precise targeted prevention and control of lake flooding risk, effectively blocking the formation of anaerobic environment while reducing energy consumption. Attached Figure Description

[0049] Figure 1 This is a schematic diagram of a method for preventing and treating lacustrine flooding based on hydrodynamic regulation, provided in an embodiment of this application.

[0050] Figure 2 This is a schematic diagram of the process for constructing a comprehensive risk assessment index for flooding of a target lake, provided in an embodiment of this application.

[0051] Figure 3 This is a schematic diagram of the flow-induced risk propagation correction process provided in the embodiments of this application.

[0052] Figure 4 This is a schematic diagram of the process for determining the instantaneous dynamic risk status of high-risk areas for lacustrine flooding based on real-time monitoring data, provided in an embodiment of this application. Detailed Implementation

[0053] Example 1: A method for preventing and controlling lacustrine flooding based on hydrodynamic regulation is provided, such as... Figure 1 As shown, the method includes the following steps:

[0054] Step 101: Obtain multi-source environmental data for the target lake.

[0055] Specifically, multi-source environmental data refers to multi-dimensional datasets that reflect the state of the lake's water environment, meteorological conditions, and sediment characteristics. Taking a shallow lake as an example, the acquired data may include water quality monitoring data, hydrological and meteorological data, sediment monitoring data, and satellite remote sensing monitoring data.

[0056] The water quality monitoring data includes indicators such as dissolved oxygen (DO), chlorophyll a, sulfides, and nitrogen and phosphorus content; the hydrological and meteorological data includes water temperature, wind speed, water level, and current velocity; and the sediment monitoring data includes the organic matter content and sulfur and iron content in the sediment. These data constitute the basic input for subsequent risk assessment. By collecting historical data from the past 5 to 10 years, data support can be provided for the long-term evolution of lake flood risk.

[0057] Step 102: Construct a comprehensive risk assessment index for lake flooding of the target lake based on multi-source environmental data, and define the spatial risk gradient distribution of high-incidence areas of lake flooding in the target lake based on the comprehensive risk assessment index for lake flooding.

[0058] In this embodiment, the comprehensive risk assessment index for lacustrine flooding is a comprehensive index that quantitatively characterizes the probability of lacustrine flooding occurring in various areas of a lake. Specifically, the target lake is spatially gridded, for example, divided into 500m × 500m grid cells, and the risk distribution of each grid cell over a long period is calculated.

[0059] The delineation process involves normalizing and weighting various risk factors. By setting preset thresholds for evaluation indicators, grid sets exceeding these thresholds are identified as high-risk areas for lacustrine flooding. The spatial risk gradient distribution describes the spatial differences in risk scores within high-risk areas; for example, the core area has a higher risk score, while the peripheral area has a lower risk score. This provides accurate spatial navigation for the non-uniform deployment of subsequent propulsion devices, avoiding the resource waste caused by blindly deploying equipment in traditional control methods.

[0060] Step 103: Using a pre-constructed three-dimensional hydrodynamic-aquatic ecology coupled model of the target lake, and constrained by a preset safe threshold for dissolved oxygen concentration, the target hydrodynamic control parameters required to bring the dissolved oxygen concentration in the high-risk lake flooding area to the safe threshold are reverse-engineered.

[0061] The three-dimensional hydrodynamic-aquatic ecosystem coupling model can be constructed based on the environmental fluid dynamics model EFDC or the three-dimensional hydrodynamic numerical model Delft3D. This model can simulate hydrodynamic characteristics such as water flow velocity, flow direction, and water level, as well as the migration and transformation processes of aquatic ecological indicators such as dissolved oxygen and chlorophyll a. The preset safe threshold for dissolved oxygen concentration is usually set at 2.0 mg / L.

[0062] The reverse engineering process involves simulating different combinations of thrust power (operating power) and flow field in the model to find the optimal solution that keeps the dissolved oxygen level in the high-incidence area above the safe threshold. Target hydrodynamic control parameters specifically include the average flow velocity, vertical mixing intensity, and horizontal diffusion rate required to meet the prevention and control objectives.

[0063] This step is used to transform ecological governance goals into calculable and actionable hydrodynamic control indicators, thereby enhancing the scientific validity of regulation plans.

[0064] Step 104: Based on the spatial risk gradient distribution and target hydrodynamic control parameters, determine the deployment density of the propulsion devices in the high-incidence area of ​​lacustrine flooding and the spatial deployment locations of each propulsion device.

[0065] In this step, the system maps the required density of equipment based on the strength of the risk gradient. Specifically, areas with higher risk scores are allocated a greater density of propulsion devices. The spatial locations of each propulsion device are represented by latitude and longitude coordinates (x, y).

[0066] For example, in the core risk grid unit, the deployment density can be set to 4-6 flow propulsion devices per square kilometer; while in the edge risk grid unit, the deployment density can be reduced to 1-2 devices per square kilometer. Based on the target hydrodynamic control parameters' requirements for the overall flow velocity, the specific location of each device is determined to ensure that the induced flow fields generated by each device can produce a synergistic effect, eliminating hydrodynamic dead zones in high-risk areas.

[0067] After determining the spatial layout of each propulsion device, the system, based on the target average flow velocity in the target hydrodynamic control parameters and the water depth and service area of ​​each spatial grid unit, combined with the hydraulic performance curves of the propulsion devices, calculates the steady-state output power required by each propulsion device to maintain the target flow velocity, and uses this power as the reference operating power P of the propulsion device. base_i .

[0068] Step 105: Obtain real-time monitoring data of the high-risk area for lacustrine flooding, determine the instantaneous dynamic risk status of the high-risk area based on the real-time monitoring data, and adjust the real-time operating power of each propulsion device according to the instantaneous dynamic risk status.

[0069] Real-time monitoring data is acquired through online monitoring sensors deployed on the lake surface. In this embodiment, the data acquisition cycle is set to 2 hours. The instantaneous dynamic risk status reflects the short-term risks caused by current weather and water quality fluctuations. The adjustment process follows the servo control principle; when the real-time monitored dissolved oxygen shows a decreasing trend or the instantaneous risk status increases, the system automatically increases the operating power of the propulsion device.

[0070] P i (t)=P base_i *R i (t);

[0071] Among them, P i (t) represents the real-time operating power of the propulsion device in space unit i at time t, P base_i R is the reference operating power of the unit's propulsion device. i (t) represents the dynamic adjustment ratio factor determined based on the instantaneous dynamic risk state. Multi-factor driven power regulation enables the system to operate at lower power to save energy during low-risk periods and at higher power to accelerate reoxygenation during high-risk periods, thus achieving a balance between prevention and control effectiveness and energy consumption.

[0072] According to another aspect of this application, a system for preventing and managing lacustrine flooding based on hydrodynamic regulation is provided, comprising:

[0073] At least one processor; and a memory communicatively connected to the at least one processor; wherein,

[0074] The memory stores instructions that can be executed by a processor to implement the steps of any one of the methods for preventing and treating lacustrine flooding based on hydrodynamic regulation proposed in this invention.

[0075] According to another aspect of this application, a computer-readable storage medium is provided, the computer-readable storage medium including a stored executable program, wherein, when the executable program is executed, it controls the device where the computer-readable storage medium is located to perform the method steps of any one of the methods for preventing and treating lacustrine flooding based on hydrodynamic regulation proposed in this invention.

[0076] Example 2: Based on Example 1 above, the specific prerequisites for constructing the comprehensive lake flood risk assessment index for the target lake are further explained in detail. In one possible implementation, the following data preprocessing steps are included before constructing the comprehensive lake flood risk assessment index:

[0077] Step 201: Based on multi-source environmental data, extract the basic risk factors of the target lake in each spatial grid cell.

[0078] Specifically, from multi-source environmental data, the following factors were extracted: water body low oxygen frequency factor, water temperature high temperature frequency factor, stable weather frequency factor, sediment endogenous risk comprehensive factor containing organic matter and sulfur and iron content, and cyanobacteria high density aggregation frequency factor detected by remote sensing images based on remote sensing monitoring data, as various basic risk factors.

[0079] After acquiring multi-source environmental data, the system spatially discretizes the target lake according to spatial grid units. Specifically, the geographic space of the target lake is divided into multiple rectangular computational grids of uniform size, such as square spatial grid units with a side length of 500m. For any spatial grid unit, the raw statistics of the above five risk factors are extracted within its corresponding historical time window.

[0080] The water hypoxia frequency factor characterizes the frequency of water hypoxia, specifically calculated as the ratio of days with measured dissolved oxygen concentrations less than or equal to the safe threshold within the spatial grid cell to the total number of monitoring days. The high water temperature frequency factor characterizes thermodynamic catalytic conditions, calculated as the percentage of days with measured water temperatures greater than 20℃. The stable weather frequency factor characterizes the state of suppressed vertical mixing in water bodies, calculated as the percentage of days with lake surface wind speeds less than 3 m / s.

[0081] The comprehensive risk factor of sediment endogenous sources, including organic matter and sulfur and iron content, reflects the material basis for the release of blackening and odor-causing substances from the sediment. The system extracts the total organic matter and sulfur and iron content data from the bottom sediment samples of each spatial grid unit. After range normalization of both, the data are linearly weighted and merged according to a preset weight ratio to obtain the original value of the comprehensive factor. For example, the weights of organic matter content and sulfur and iron content are each set to 0.5.

[0082] The high-density aggregation frequency factor of cyanobacteria detected by remote sensing imagery is used to characterize the source risk of mass phytoplankton death and decomposition. The concentration of chlorophyll a in the water body is extracted by using high-resolution satellite imagery of the target area, and the proportion of high-density detections with chlorophyll a concentration greater than 100 μg / L is calculated.

[0083] In some optional implementations, when long-term cloud cover over the target area leads to missing satellite remote sensing image data, chlorophyll a inversion data of the corresponding water area can be obtained using a multispectral camera mounted on a drone for replacement extraction. Furthermore, for the extraction of comprehensive factors of endogenous risk in sediment, sediment thickness can be introduced as a third-dimensional weighted feature, and the assessment results of endogenous risk in sediment can be further corrected by evaluating the total sediment load.

[0084] Step 202: Normalize each basic risk factor and uniformly map it into a dimensionless risk score.

[0085] Since the extracted basic risk factors have different dimensions and numerical distribution ranges, in order to achieve equivalent mathematical coupling of multiple factors in subsequent spatial assessment, normalization is required to map them to the same numerical range. In this embodiment, the system uses the range normalization method to perform a linear transformation on each basic risk factor, mapping it to a dimensionless numerical space of 0 to 1.

[0086] f i_k =(F i_k -F k_min ) / (F k_max -F k_min );

[0087] Among them, f i_k F is the normalized risk score obtained by processing the k-th basic risk factor in spatial grid cell i. i_k F represents the raw statistical value extracted from the k-th basic risk factor in spatial grid cell i. k_min F represents the historical minimum statistical value of the k-th basic risk factor across all spatial grid cells in the entire target lake area. k_max This represents the historical maximum statistical value of the k-th basic risk factor across all spatial grid cells in the entire target lake area.

[0088] Through the above calculations, all data with different dimensions are uniformly defined as risk scores. Assuming the original statistical value of the high-temperature frequency factor extracted from a certain spatial grid cell is 30 days, the historical maximum value is 50 days, and the historical minimum value is 10 days, substituting these values ​​into the above formula yields the normalized score of the high-temperature frequency factor for this spatial grid cell: (30-10) / (50-10) = 0.5. This dimensionless score represents the relative level of this grid cell in the global high-temperature risk.

[0089] In some alternative implementations, when the raw data of a certain basic risk factor exhibits a significant non-uniform long-tail distribution, such as the low frequency of high-density cyanobacterial aggregation in most grids and only high frequency in a few lake estuary grids, using conventional range normalization will result in the scores of a large number of data points being compressed to close to 0.

[0090] For such situations, logarithmic transformation normalization can be used instead of range normalization. Specifically, a base-10 logarithmic operation is performed on the original statistical values ​​to compress extreme differences in high-value ranges, and then the maximum and minimum values ​​of the transformed data are extracted for range mapping. Through pre-processing with logarithmic transformation, the risk discrimination of intermediate range data can be preserved, avoiding distortion and interference from extreme outliers on the overall risk distribution assessment.

[0091] According to one aspect of this application, a comprehensive lake flood risk assessment index is constructed based on multi-source environmental data, including the following data preprocessing steps:

[0092] The target lake is divided into spatial grids to obtain each spatial grid unit;

[0093] Based on multi-source environmental data, we extract the basic risk factors of the target lake in each spatial grid cell.

[0094] The basic risk factors are normalized and uniformly mapped to dimensionless risk scores.

[0095] Example 3 further details the nonlinear lacustrine flooding risk coupling assessment process and the corresponding interaction pair screening mechanism. Lacustrine flooding is usually the result of the synergistic effect of multiple risk factors. For example, high temperature or low oxygen alone may not trigger flooding; however, when both co-occur, the probability of an outbreak increases nonlinearly. This example quantifies this synergistic amplification effect by introducing a second-order interaction term.

[0096] Construct comprehensive alpine flood risk assessment indicators for the target lake, such as... Figure 2 As shown, the process includes the following:

[0097] For any spatial grid cell, based on its normalized basic risk factors, the main effect risk contribution of each individual risk factor when acting independently is calculated; pre-screened risk factor pairs with significant positive synergistic amplification effects are obtained, and the synergistic interaction risk contribution of the risk factor pair is calculated based on the product of the scores when the two basic risk factors co-occur; based on the preset interaction intensity adjustment parameter, the main effect risk contribution and the synergistic interaction risk contribution are nonlinearly weighted and fused to obtain the comprehensive risk assessment index for lake flooding.

[0098] Specifically, for each spatial grid cell i, the system calculates the weighted sum of the main effects of its five normalized risk factors. The main effect risk contribution reflects the independent contribution of each factor to the occurrence of lake flooding.

[0099] Based on this, using a pre-determined set of significant interaction factor pairs S, the product of the factor scores for each pair in the set is extracted, multiplied by the corresponding interaction weight, and then summed to obtain the collaborative interaction risk contribution. The system uses the interaction strength adjustment parameter λ to nonlinearly weight and fuse the two contributions. The formula is as follows:

[0100] LBRI i_NL =(1 / (1+λ))*(Σ(w k *f i_k )+λ*Σ(w kl *f i_k *f i_l ));

[0101] Among them, LBRI i_NL Σ is the nonlinear comprehensive risk assessment index for lake flooding of spatial grid cell i; λ is the interaction intensity adjustment parameter, used to control the contribution ratio of the interaction term to the main effect term, and can be selected as 0.5; Σ is the summation operator; w k f is the main effect weight of the k-th risk factor; i_k w is the normalized score of the k-th risk factor in spatial grid cell i; kl f represents the co-interaction weights of risk factor pairs (k, l) in set S; i_l This is the normalized score of the l-th risk factor in spatial grid cell i.

[0102] For example, for a large, shallow lake, λ can be set to 0.5. If the hypoxia factor score of a certain grid cell is 0.8 and the water temperature factor score is 0.7, the two are identified as a significant positive interaction pair. In the traditional linear model, the weighted contribution of the two is only the sum of their independent contributions; however, in the nonlinear model of this embodiment, an additional interaction contribution term of 0.8 * 0.7 = 0.56 is added, which more accurately reflects the sharp increase in the risk of lacustrine flooding under high temperature and hypoxia conditions.

[0103] In this embodiment, the main effect weights w of each basic risk factor are... k The entropy weight method is used to determine the main effect weights of each factor. By calculating the information entropy of each factor across all spatial grid cells, and using the information entropy to reflect the amount of information carried by the factor, the main effect weights of each factor are objectively determined.

[0104] Specifically, the information entropy E for calculating the normalized score of the k-th risk factor across all spatial grid cells is calculated. k :

[0105] Ek =-(1 / ln(N))×Σ(p i_k ×ln(p i_k ));

[0106] Where, p i_k Let N be the proportion of the k-th factor score of spatial grid cell i to the total score of that factor across the entire domain, and N be the total number of grid cells.

[0107] Further calculate the difference coefficient d for each factor. k =1-E k and d k Normalization yields w k =d k / Σ(d j ), where j is the summation index, representing all risk factors involved in the evaluation.

[0108] In some alternative implementations, the main effect weights can also be determined by domain experts based on their experience with the historical ecological characteristics of the target lake.

[0109] In some embodiments, before obtaining pre-screened risk factor pairs with significant positive synergistic amplification effects, an offline construction step is included to determine the risk factor pairs:

[0110] From the pre-stored historical monitoring database, the actual lake flooding status identifiers for each time period and each spatial grid unit within the historical monitoring period are extracted. Using the actual lake flooding status identifiers as the dependent variable and each basic risk factor and all its second-order interaction combination terms as independent variables, a historical statistical regression model is established. The significance of each second-order interaction combination term in the historical statistical regression model is tested, and the interaction terms that have statistical significance and positively amplify the probability of lake flooding are retained as pre-screened risk factor pairs with significant positive synergistic amplification effects.

[0111] This step is part of the offline construction process and is used to determine the set S and interaction weights w involved in the preceding steps. kl Specifically, the system retrieves historical monitoring databases from the past 10 years, defining the state identifier y as whether lake flooding actually occurred at each sample point during that period, with y=1 indicating occurrence and y=0 indicating no occurrence. A logistic regression model is established using five basic factors and their combinations, generating 10 second-order interaction terms as independent variables.

[0112] ln(Pr / (1-Pr))=α0+Σ(α k *f k )+Σ(α kl *f k *f l );

[0113] Where Pr is the predicted probability of lake flooding; f k f is the normalized score of the k-th risk factor; l α is the normalized score of the l-th risk factor; α0 is the intercept term; α k The regression coefficient of the main effect term; α kl The regression coefficients are the interaction terms.

[0114] The Wald test was used to evaluate the coefficient α of each interaction term. kl Perform a significance test. If the significance level of a certain interaction term is p < 0.05, and the coefficient α kl If the value is greater than 0, then the factor is considered to have a significant positive synergistic amplification effect and is included in set S. The corresponding interaction weight w kl It is obtained by normalizing the absolute values ​​of all significant interaction coefficients in set S.

[0115] In some alternative implementations, third-order interaction terms can be introduced into the independent variables of the regression model to capture more complex ecological linkages. Furthermore, the significance test can be replaced by the likelihood ratio test instead of the Wald test to improve the test power with small samples of historical data. Parameter calibration methods driven by historical data can ensure that the risk assessment logic is highly adapted to the predetermined ecological mechanisms of the target lake.

[0116] Example 4 further details the flow-induced risk propagation correction process. Because water flow in shallow lakes leads to the spatial transport of risk factors, the actual risk in downstream areas is often influenced by upstream inputs. This example provides a risk propagation correction method based on hydrodynamic connectivity.

[0117] like Figure 3 As shown, after obtaining the comprehensive risk assessment index for lacustrine flooding, and before defining the spatial risk gradient distribution of high-incidence areas of lacustrine flooding in the target lake, a flow-induced risk propagation correction step is also included:

[0118] Step 401: Based on the pre-constructed three-dimensional hydrodynamic-aquatic ecology coupled model of the target lake, extract the baseline net flux of water flow between adjacent spatial grid cells in the target lake, and determine the upstream connected neighborhood of each spatial grid cell based on the baseline net flux of water flow.

[0119] In this embodiment, the system invokes a pre-built three-dimensional hydrodynamic-aquatic ecosystem coupling model to extract the time-averaged flow field for the target time period. Specifically, the three-dimensional hydrodynamic-aquatic ecosystem coupling model can be constructed using the EFDC model or the Delft3D model. The extracted flow field is the baseline flow field under natural driving conditions, i.e., the state that does not include the artificial hydrodynamic increments generated by the operation of the propulsion device.

[0120] For any two adjacent spatial grid cells sharing a boundary, the system calculates the time-averaged water flow through that shared boundary as the baseline net flux. If the baseline net flux from the adjacent spatial grid cell to the local spatial grid cell is greater than 0, then the adjacent spatial grid cell is determined to belong to the upstream connected neighborhood of the local spatial grid cell. The set of all spatial grid cells satisfying this condition is defined as the complete upstream connected neighborhood of the local spatial grid cell. The total flux flowing into the local spatial grid cell is the sum of all baseline net fluxes within the upstream connected neighborhood.

[0121] Step 402: Based on the comprehensive risk assessment index of lake flooding of each spatial grid unit, according to the preset connectivity influence coefficient, the risk increment of each unit in the upstream connected neighborhood after flux ratio weighting is iteratively accumulated to the local spatial grid unit until the iteration converges, and the final comprehensive risk assessment index of lake flooding after spatial propagation correction is obtained.

[0122] In some embodiments, this step may also be to use the comprehensive risk assessment index of each spatial grid cell as the initial iteration value, and according to the preset connectivity influence coefficient, to iteratively accumulate the risk increment of each cell in the upstream connected neighborhood after flux ratio weighting to the local spatial grid cell until the maximum difference in the whole domain between two adjacent iterations is less than the preset convergence threshold, so as to obtain the final comprehensive risk assessment index of the lake after spatial propagation correction.

[0123] Specifically, the system draws on the transition probability concept of Markov chains to construct iterative correction logic. The actual risk of a local spatial grid unit is composed of the basic risk arising from local natural evolution and the propagation risk brought by upstream water inflow. The system uses the comprehensive lake flooding risk assessment index obtained in Example 3 as the input value for the 0th iteration. In a single iteration, the system first calculates the product of the risk score and its flux ratio for each unit in the upstream connected neighborhood, sums all products to obtain the propagation risk increment, and uses the connectivity influence coefficient to linearly combine the local basic risk and the propagation risk increment. The above iterative process is implemented through the following formula:

[0124] LBRI SC_i (n+1)=(1-β)*LBRI i +β*Σ((Q ji / Q total_i )*LBRI SC_j (n));

[0125] Among them, LBRI SC_i (n+1) represents the comprehensive risk assessment index for lake flooding of local spatial grid cell i after spatial propagation correction following the (n+1)th iteration, where β is the preset connectivity influence coefficient, and LBRI iLet Σ be the comprehensive risk assessment index for lake flooding of local spatial grid cell i, and let Q be the summation operator for all upstream connected neighboring spatial grid cells j. ji Q is the baseline net flux of water flow from spatial grid cell j to local spatial grid cell i. total_i For the total flux flowing into local spatial grid cell i, when Q total_i When the value is zero, the incremental term for transmission risk is counted as zero, and the LBRI (Less than 100% LBRI) is zero. SC_j (n) is the comprehensive risk assessment index of lake flooding after spatial propagation correction of spatial grid cell j after the nth iteration.

[0126] In the above formula, the connectivity influence coefficient β controls the correction magnitude of the upstream risk input on the local evaluation results. In this embodiment, the connectivity influence coefficient β is set to 0.2. The system continuously executes the above iterative calculation. After each iteration, it calculates the absolute value of the difference in risk indicators between two adjacent iterations for all spatial grid cells. When the absolute value of the largest difference across the entire domain is less than 0.0001, the system determines that the iteration has converged and stops the calculation. The index array output at this time is the final comprehensive risk assessment index for lake flooding.

[0127] Through this iterative mechanism, risk signals in the water backflow area are repeatedly accumulated, reflecting the true high-risk state of areas with poor water exchange.

[0128] Step 403: Based on the final comprehensive risk assessment index for lacustrine flooding, spatial clustering is performed to determine the continuous boundary of the high-incidence area of ​​lacustrine flooding and its spatial risk gradient distribution.

[0129] To transform scattered high-risk grid cells into a continuous, engineering-feasible prevention and control area, the system needs to remove isolated noisy cells. All spatial grid cells whose final comprehensive lake flood risk assessment index is greater than or equal to a preset risk benchmark threshold are extracted. The DBSCAN algorithm, a density-based spatial clustering algorithm with noise, is used to cluster the spatial grid cells. Optionally, the neighborhood search radius is set to 1-2 grid intervals, and the minimum number of grid cells for core points is set to 3-5.

[0130] The preset risk baseline value can be determined by ROC curve analysis on a validation dataset based on the historical frequency of flooding in the target lake and the required precision of prevention and control. In this embodiment, the preset risk baseline value is set to 0.5.

[0131] Specifically, the system sets the neighborhood search radius and the minimum number of grid cells required to form the core point. The neighborhood search radius and the minimum number of grid cells can be determined using conventional parameter calibration methods based on the size of the spatial grid cells and the spatial characteristics of the target lake.

[0132] The outer contour corresponding to the largest connected component identified by this clustering algorithm is thus determined as the continuous boundary of the high-risk lacustrine flooding area. The set of risk assessment indicators for each spatial grid cell located within this continuous boundary directly constitutes the spatial risk gradient distribution. This gradient distribution data will be used as subsequent input to quantitatively map the deployment density of the propulsion device.

[0133] Example 5: Detailed explanation of the hydrodynamic demand back-engineering process based on a coupled model. The back-engineering uses numerical simulation to establish a quantitative functional relationship between hydrodynamic intensity and dissolved oxygen response, determining the optimal energy consumption control index while ensuring environmental safety.

[0134] Using a pre-constructed three-dimensional hydrodynamic-aquatic ecosystem coupled model of the target lake, and constrained by a preset safe dissolved oxygen concentration threshold, the target hydrodynamic regulation parameters required to bring the dissolved oxygen concentration in the high-risk lake flooding area to the safe threshold are derived in reverse, including:

[0135] Multiple pre-set hydrodynamic settings of different intensities are input into the three-dimensional hydrodynamic-aquatic ecology coupled model of the target lake; the spatiotemporal distribution of dissolved oxygen in the lake flood-prone area is simulated under each set of hydrodynamic settings; the minimum average water velocity, vertical mixing intensity and horizontal diffusion rate combination required to keep the dissolved oxygen in the entire lake flood-prone area stable above the safe threshold of dissolved oxygen concentration are extracted as the target hydrodynamic regulation parameters.

[0136] Specifically, the three-dimensional hydrodynamic-aquatic ecosystem coupled model pre-defines the grid and calibrates the parameters based on the topography and hydrological characteristics of the target lake. Taking a high-incidence area of ​​a lake as an example, the model uses an orthogonal curved grid, divided into 5 to 10 layers in the vertical direction to accurately simulate water stratification. Before performing backpropagation, the system first determines the boundary conditions of the model input, including wind field, inflow rate, and sediment oxygen consumption rate.

[0137] For the high-risk lacustrine flooding areas defined in step 102, the system sets a set of hydrodynamic intensity gradient matrices as disturbance inputs. The gradient matrix is ​​constructed by dividing the range from the estimated minimum effective flow velocity to the rated maximum flow velocity of the propulsion device into several steps at equal or logarithmic intervals. For example, the average flow velocity of the water body is set from 0.01 m / s to 0.20 m / s, and 20 gradient conditions are constructed with a step size of 0.01 m / s.

[0138] The hydrodynamic settings specifically include different thrust flow gradients and different turbulent diffusion coefficients. The dissolved oxygen concentration evolution process in each grid cell within the high-incidence area is calculated in parallel using the model under each set of operating conditions. In this embodiment, the target simulation period is the summer high-temperature stable period, for example, simulating continuously for 15 to 30 days. Engineers can select an appropriate simulation period based on the seasonal characteristics of high-incidence lake flooding in the target lake.

[0139] After obtaining simulation results for multiple operating conditions, the system retrieves the spatial minimum value of dissolved oxygen concentration across the entire high-incidence area under each operating condition. It then extracts a set of operating conditions that consistently maintain a dissolved oxygen concentration greater than or equal to the safe threshold. Within this set of operating conditions, it searches for the set of parameter values ​​that result in the lowest energy consumption contribution or the lowest mechanical strength.

[0140] v target =min(v i |DO min_i ≥DO limit );

[0141] Among them, v target For the target average flow velocity in the determined target hydrodynamic control parameters, v i For the set flow rate intensity in the i-th simulation condition, DO min_i Let DO be the spatiotemporal minimum value of dissolved oxygen concentration in the high-incidence area under the i-th simulated operating condition. limit This is the preset safe threshold for dissolved oxygen concentration.

[0142] Similarly, the system simultaneously extracts the corresponding minimum vertical mixing intensity and minimum horizontal diffusion rate. Vertical mixing intensity is typically characterized by the vertical eddy viscosity coefficient or vertical diffusion coefficient, and is a technical quantity used to evaluate the vertical exchange efficiency of water bodies. Through back-calculation, the system can identify the critical kinetic conditions for inhibiting anaerobic reduction reactions and blocking sulfide formation.

[0143] According to one aspect of this application, similarly, the system satisfies DO min_i ≥DO limit Within the constrained set of operating conditions, the minimum vertical mixing intensity and minimum horizontal diffusion rate are extracted simultaneously. When the optimal values ​​of the three parameters do not appear in the same set of operating conditions, the system selects the operating condition with the minimum comprehensive energy consumption that simultaneously satisfies the constraints of the three parameters as the target hydrodynamic control parameter.

[0144] In some optional implementations, the system can perform the aforementioned reverse inference for different seasons or meteorological scenarios, such as typical extreme high-temperature scenarios and long-period stable weather scenarios, to construct a target hydrodynamic control parameter library under multiple scenarios. During actual control, the target hydrodynamic control parameters that are closest to the current meteorological driving background can be matched from the parameter library based on short-term weather forecast results, thereby improving the predictability and accuracy of control.

[0145] Example 6 describes the specific process of determining the deployment density of propulsion devices in areas prone to lacustrine flooding and the spatial deployment locations of each device. In one possible implementation, the process includes the following steps:

[0146] Step 601: Based on the spatial risk gradient distribution and target hydrodynamic control parameters, determine the deployment density of the propulsion device in the high-incidence area of ​​lacustrine flooding, including: determining the minimum and maximum deployment density benchmarks required to maintain the safe threshold of dissolved oxygen concentration based on the target hydrodynamic control parameters; obtaining a pre-configured control morphology index, which is used to control the nonlinear change trend of equipment deployment density with risk level (risk score); establishing a monotonically increasing mapping relationship from regional risk score to equipment demand density using the minimum deployment density benchmark, the maximum deployment density benchmark, and the control morphology index; substituting the risk score of each spatial grid unit in the spatial risk gradient distribution into the mapping relationship to calculate the target demand density of each spatial grid unit, which serves as the quantitative benchmark for deployment density.

[0147] After obtaining the spatial risk gradient distribution, the system needs to transform the abstract dimensionless risk score into engineering layout parameters for physical equipment. Specifically, based on the hydrodynamic back-engineering results, the minimum number of equipment per unit area required to maintain dissolved oxygen levels at the boundary of high-risk areas is extracted and defined as the minimum deployment density benchmark.

[0148] The determination of the aforementioned density benchmark also incorporates the individual performance parameters of the selected propulsion devices, including the effective propulsion flow rate generated by a single device at rated power and the effective radius of action in the horizontal direction. Based on this, the system calculates the minimum and maximum number of devices per unit area required to maintain the target hydrodynamic control parameters at the high-risk zone boundary and the core highest-risk zone, respectively. The benchmark operating power of each propulsion device is set to its rated operating power level required to maintain the target hydrodynamic control parameters within the corresponding spatial grid cell. Engineers can determine these parameters based on the performance specifications of the actual selected equipment.

[0149] Similarly, the maximum number of devices per unit area required to maintain dissolved oxygen levels within the highest-risk core area is extracted and defined as the benchmark for maximum deployment density. The system constructs a mapping function by introducing a regulation morphology index. This mapping function is based on the interval formed by the risk benchmark value and the maximum risk extreme value, and the calculation process is as follows:

[0150] ρ i_d =ρ min +(ρ max -ρ min )*((LBRI SC_i -LBRI th ) / (LBRI max -LBRI th )) γ ;

[0151] Where, ρ i_d Let ρ be the target required density of spatial grid cell i. min As the minimum deployment density benchmark, ρmax LBRI is the benchmark for the highest deployment density. SC_i LBRI is the local risk score for spatial grid cell i. th To define the risk benchmark threshold for high-incidence areas, LBRI max γ represents the maximum risk score within high-incidence areas, and γ is the regulatory pattern index.

[0152] In the formula, the regulation morphology index directly determines the spatial tilt strategy of resource allocation. When γ > 1, the mapping curve exhibits a downward convex shape, indicating that the deployment density increases slowly at the low-risk end and rapidly at the high-risk end. Under this configuration, equipment resources will be highly concentrated in the core high-risk area. When 0 < γ < 1, the mapping curve exhibits an upward convex shape, indicating that the system tends to achieve balanced equipment coverage throughout the high-risk area. Taking the summer regulation of a certain lake as an example, γ is set to 2.0 to ensure that limited power resources are concentrated in the local core waters prone to lake flooding. Engineers can make adaptive adjustments by comparing the overall dissolved oxygen compliance rate under different γ values ​​in the simulation model, based on the high-risk area of ​​the target lake, the total amount of available equipment, and the control precision requirements.

[0153] In some alternative implementations, when the target lake has a uniform water depth and a flat lakebed, the differences in hydrodynamic response sensitivity among the various spatial grid cells are small. In this case, the system can set the control morphology index to γ=1.0, so that the target demand density and risk score have a linear mapping relationship, thereby simplifying the computational complexity.

[0154] Step 602, determine the spatial layout locations of each propulsion device, specifically including:

[0155] Obtain the area of ​​each spatial grid cell in the high-incidence area of ​​lacustrine flooding and the total number of pre-set propulsion devices available; calculate the theoretical number of units required for each spatial grid cell based on the target demand density and corresponding area of ​​each spatial grid cell.

[0156] Under the resource constraint that the total number of available propulsion devices X is less than the total number of theoretically required devices Y, i.e. when X < Y, an allocation optimization criterion is constructed with the goal of minimizing the global area-weighted residual risk. The residual risk of a single grid cell is determined by the ratio of the difference between the actual number of allocated devices and the theoretically required number of devices, as well as the local risk score. According to the allocation optimization criterion, propulsion devices are preferentially allocated to the spatial grid cells with the highest risk elimination efficiency per unit of equipment input until the allocation reaches the total available number limit, thereby determining the actual number of devices deployed in each spatial grid cell and the spatial deployment points of each propulsion device within it.

[0157] Furthermore, when X≥Y, the theoretical number of units required for each spatial grid unit is directly used as the actual number of units deployed, and the spatial deployment points of each propulsion device within each spatial grid unit are determined accordingly.

[0158] Because there is a physical upper limit to the procurement and stockpiling of propulsion devices in actual engineering, the system must handle the resource constraint problem where the target demand exceeds the available inventory. First, the theoretical demand number of devices for each spatial grid cell is calculated. The target demand density is multiplied by the area of ​​that spatial grid cell, and the product is rounded up. The total available number of propulsion devices in the current inventory is obtained. When the total available number is less than the sum of the theoretical demand numbers for all grids, the system constructs an integer programming model for resource scheduling. The objective function of the model is constructed as follows:

[0159] Z obj =Σ(LBRI SC_i *max(0, 1-m) i / m i_star )*A i );

[0160] Among them, Z obj Σ represents the total area-weighted residual risk, where Σ is the summation operator for all spatial grid cells in high-incidence areas, and LBRI is the summation operator. SC_i Let m be the local risk score of spatial grid cell i, and max be the function to maximize the value. i m represents the actual number of units deployed in spatial grid cell i. i_star Let A be the theoretical number of units required for spatial grid cell i. i Let i be the area of ​​the spatial grid cell i.

[0161] To solve the aforementioned optimization problem with resource upper limit constraints, the system employs a greedy allocation algorithm for per-device scheduling. Specifically, the risk reduction contribution per unit device in each spatial grid cell is calculated, and its value is equal to the product of the local risk score and the corresponding area, divided by the theoretically required number of devices.

[0162] All spatial grid cells are sorted in descending order of elimination efficiency. In each allocation cycle, the system allocates one device to the top-ranked spatial grid cell, updates the actual number of devices deployed and the remaining gap in that cell, and recalculates the ranking until the total number of devices allocated reaches the total available quantity. After determining the actual number of devices deployed, the system combines local micro-topography and water depth data within each spatial grid cell to evenly deploy the devices at the geometric center of the grid or along the main water flow channels, outputting the specific latitude and longitude coordinates as the final spatial deployment points.

[0163] Furthermore, if the system detects that X≥Y, i.e., the resources are sufficient, the system will directly skip the integer programming allocation criteria and set the actual number of units deployed according to the theoretical number required for each spatial grid unit, thereby shortening the calculation time and accelerating the project deployment.

[0164] Example 7 describes the specific process of determining the instantaneous dynamic risk status of high-risk areas for lacustrine flooding based on real-time monitoring data, as well as the offline calibration process of the dynamic calculation weights of pre-configured monitoring indicators. While the static evaluation in the preceding examples primarily addresses the spatial planning problem of equipment deployment, the two-layer dynamic evaluation architecture constructed in this example addresses the time-based decision-making problem of real-time equipment power regulation.

[0165] In one possible implementation, the instantaneous dynamic risk status of areas prone to lacustrine flooding is determined based on real-time monitoring data, such as... Figure 4 As shown, it includes the following steps:

[0166] Step 701: Obtain the preset absolute safety threshold and preset extreme danger threshold corresponding to each monitoring indicator in the real-time monitoring data.

[0167] Real-time monitoring data includes continuous physical quantities such as dissolved oxygen concentration, water temperature, wind speed, and chlorophyll a concentration within the current spatial grid cell. The preset absolute safety threshold is defined as the boundary where the physical quantity is at a level that completely eliminates the risk of inducing lake flooding, while the preset extreme danger threshold is defined as the limit where the physical quantity is at a level that easily triggers the deterioration of the water body into a black and odorous state.

[0168] Taking dissolved oxygen concentration as an example, the preset absolute safety threshold can be set at 2.0 mg / L, which indicates sufficient oxygen in the water. The preset extreme danger threshold can be set at 0.5 mg / L, which indicates that an anaerobic environment has formed. Taking water temperature as an example, the preset absolute safety threshold is set at 20℃, and the preset extreme danger threshold is set at 30℃. For different types of lakes, these thresholds can be specifically calibrated by analyzing hydrological characteristics and basic ecological data.

[0169] Step 702: Using segmented risk conversion logic, the real-time physical measurement values ​​of each monitoring indicator are mapped to dimensionless instantaneous risk scores. Measurement values ​​at or above the preset absolute safety threshold are mapped to zero risk scores, and measurement values ​​at or below the preset extreme danger threshold are mapped to full risk scores. Measurement values ​​between the preset absolute safety threshold and the preset extreme danger threshold are linearly interpolated according to their relative positions between the two thresholds to obtain instantaneous risk scores between zero risk scores and full risk scores.

[0170] This step converts physical measurements of different dimensions into instantaneous risk scores between 0 and 1 with a uniform metric. When a measured value falls between a preset absolute safety threshold and a preset extreme danger threshold, linear interpolation is used to calculate its corresponding score. For negatively correlated indicators such as dissolved oxygen, where lower values ​​indicate higher risk, the calculation formula for the segmented risk conversion logic is as follows:

[0171] g1=(DO limit -DO real ) / (DO limit -DO crit );

[0172] Where g1 is the instantaneous risk score of dissolved oxygen concentration, DO limit To preset an absolute safety threshold, DO crit To preset extreme danger thresholds, DO real For dissolved oxygen falling between these two values, the real-time physical measurement is used. For dissolved oxygen at or above DO... limit For measured values, i.e., values ​​greater than or equal to 2.0 mg / L, g1 is directly assigned a value of 0; for values ​​at or below DO... crit For measured values, i.e., values ​​less than or equal to 0.5 mg / L, g1 is directly assigned a value of 1. For positively correlated indicators such as water temperature, where higher values ​​indicate greater risk, the numerator is replaced with the real-time measured value minus the preset absolute safety threshold, and the denominator is replaced with the preset extreme danger threshold minus the preset absolute safety threshold. Since endogenous risk indicators in sediment remain stable in the short term, their real-time physical measurements can directly use the static normalized score obtained from the most recent monitoring sampling.

[0173] According to one aspect of this application, the dynamic calculation weights of the pre-configured monitoring indicators are pre-calibrated through the following offline construction steps a to c:

[0174] Step a: Extract historical monitoring data from the target lake's historical records within a preset time window preceding each actual lake flooding event. In other words, extract historical monitoring data from the pre-stored historical monitoring database within a preset time window preceding each actual lake flooding event of the target lake.

[0175] Static assessments using spatial distribution information entropy cannot reflect the sensitivity of various factors to triggering lake flooding over time. Therefore, a time-series-based conditional probability analysis mechanism needs to be constructed. The system retrieves historical records of the target, filters out all recorded actual lake flooding events, and extracts historical monitoring data within a preset time window preceding each event. Specifically, the preset time window can be set to 24 hours before the event occurs.

[0176] Step b: Based on the segmented risk conversion logic, calculate the average historical risk score of each monitoring indicator within a preset time window before each actual lake flooding event.

[0177] The system employs the same segmented risk conversion logic as step 702, mapping the extracted historical monitoring data one by one to a risk score between 0 and 1. For each monitoring indicator, its average risk score within a preset time window prior to all previous events is calculated. This average risk score reflects the average level of high risk of various physical indicators during the incubation period before an impending lacustrine flooding disaster.

[0178] Step c: Determine the dynamic calculation weight of each pre-configured monitoring indicator according to the proportion of the historical risk score average of each monitoring indicator in the total average score of all indicators, so that the monitoring indicators that showed a higher average risk status before the historical lake flooding outbreaks will receive higher calculation weights.

[0179] The ratio of the historical average risk score of each monitoring indicator to the sum of the average scores of all indicators is normalized to form a dynamic weight.

[0180] w k_D =g k_mean / Σ(g j_mean );

[0181] Among them, w k_D For the dynamic calculation of the k-th monitoring indicator, g k_mean Let g be the historical risk score average of the k-th monitoring indicator, Σ be the summation operator for all monitoring indicators j, and g be the summation operator for all monitoring indicators j. j_mean is the historical risk score average of the j-th monitoring indicator.

[0182] For example, if historical data shows that dissolved oxygen concentrations were almost always at extremely low levels before the occurrence of lake flooding, i.e., the average historical risk score was close to 1, while the average score for wind speed was only 0.4, then dissolved oxygen would be assigned a much larger dynamic calculation weight than wind speed, reflecting its control sensitivity as a major inducing factor.

[0183] Step 703: Obtain the dynamic calculation weights of each pre-configured monitoring indicator; sum the instantaneous risk score of each monitoring indicator with the corresponding dynamic calculation weight to obtain the instantaneous dynamic risk status that characterizes the sensitivity to multi-factor comprehensive triggering.

[0184] During the real-time online operation phase, the system directly reads the dynamically calculated weights that have been calibrated offline and performs a weighted summation operation on them and the instantaneous risk scores calculated in step 702.

[0185] LBRI D =Σ(w k_D *g k );

[0186] Among them, LBRI DLet Σ be the instantaneous dynamic risk state of the spatial grid cell at the current moment, and w be the summation operator. k_D For the dynamic calculation of the k-th monitoring indicator, g k Let be the instantaneous risk score of the k-th monitoring indicator.

[0187] Assuming a grid cell has a current instantaneous dissolved oxygen risk score of 0.8 and a current instantaneous water temperature risk score of 0.9, and the pre-configured dynamic calculation weights for dissolved oxygen and water temperature are 0.6 and 0.4 respectively, then the instantaneous dynamic risk state calculation result at this moment is 0.8*0.6 + 0.9*0.4 = 0.84. This result is a continuous real number between 0 and 1, used to indicate the degree of environmental degradation and drive the power servo action of subsequent propulsion devices.

[0188] Example 8 further describes the specific process of adjusting the real-time operating power of each propulsion device. In one possible implementation, it includes the following steps:

[0189] Step 801: Obtain the pre-configured low-risk response threshold and high-risk warning threshold.

[0190] Specifically, the system needs to divide the continuous risk space into different management intervals to match different power output strategies. The low-risk response threshold is used to define the boundary when the water body is in a safe or very low-risk state, while the high-risk warning threshold is used to define the boundary when the water body enters a dangerous or very high-risk state.

[0191] In practical engineering applications, the pre-configured low-risk response threshold can be set to 0.3, and the pre-configured high-risk warning threshold can be set to 0.6. These two values ​​divide the dimensionless risk space of 0 to 1 into three warning intervals: the energy-saving maintenance interval, the routine prevention and control interval, and the emergency response interval.

[0192] Step 802: Determine the position of the instantaneous dynamic risk status within different warning intervals divided by the low-risk response threshold and the high-risk warning threshold.

[0193] The system obtains the instantaneous dynamic risk state value output in Example 7 and compares it with the two threshold values ​​obtained in step 801. When the instantaneous dynamic risk state LBRI... D (t)≤L low When the risk level reaches the low-risk response threshold, the system determines that the current risk falls within the energy-saving maintenance range. When L... low <LBRI D (t)≤L high When the (high-risk warning threshold) is reached, the current risk is determined to fall within the normal prevention and control range. When LBRI D (t)>L highAt that time, it is determined that the current risk falls within the emergency response range. By determining the range, the system can identify the current dynamic evolution stage of the aquatic environment.

[0194] Step 803: Calculate the dynamic adjustment ratio factor for smooth transition using continuous adjustment logic that matches the landing warning interval.

[0195] To avoid frequent start-ups and shutdowns of the power generation equipment and the impact on the power grid caused by traditional binary switch control, the system is configured with a segmented and continuous proportional control mapping relationship. The dynamic adjustment proportional factor is calculated as follows:

[0196] Within the energy-saving maintenance range, i.e., when LBRI D (t)≤L low :R i (t)=R min ;

[0197] Within the normal prevention and control zone, that is, when L low <LBRI D (t)≤L high :

[0198] R i (t)=R min +(1-R min )*(LBRI D (t)-L low ) / (L high -L low );

[0199] Within the emergency response interval, i.e., when LBRI D (t)>L high :

[0200] R i (t)=1+(R max -1)*(LBRI D (t)-L high ) / (1-L high );

[0201] Among them, R i (t) is the dynamic adjustment scaling factor of spatial grid cell i at time t, R min R is the minimum adjustment ratio. max For the highest adjustment ratio, LBRI D (t) represents the instantaneous dynamic risk state at time t, L low As a low-risk response threshold, L high This is the high-risk warning threshold.

[0202] In the above logic, the minimum regulation ratio is used to maintain a slight flow of the bottom water during low-risk periods, and its value can be set to 0.3. The maximum regulation ratio is used to forcibly accelerate reoxygenation before a lacustrine flooding event, and its value can be set to 1.5.

[0203] Through the above formula, the dynamic adjustment ratio factor exhibits a mathematical characteristic of continuous and smooth growth as risk increases. Assume the instantaneous dynamic risk state of a certain spatial grid cell is 0.84, exceeding the set high-risk warning threshold of 0.6, thus falling into the emergency response range. The system calculates 1 + (1.5 - 1) * (0.84 - 0.6) / (1 - 0.6) = 1.3. This indicates that the system needs to intervene dynamically at a rate 1.3 times higher than the normal full-load baseline.

[0204] The above thresholds and minimum adjustment ratio R min and the highest adjustment ratio R max The specific value can be determined through on-site debugging or simulation optimization, based on the historical risk distribution characteristics of the target lake and the rated operating range of the propulsion device.

[0205] Step 804: Obtain the preset reference operating power of each propulsion device, and multiply the reference operating power by the dynamic adjustment ratio factor to set the real-time operating power of each propulsion device, thereby realizing continuous servo adjustment based on the fluctuation of instantaneous risk state.

[0206] The system reads the reference operating power of the propulsion device allocated within the spatial grid cell during the static deployment planning phase. It performs a multiplication operation to calculate the product of the reference operating power and the dynamic adjustment proportional factor, and sets the result of this product as the target operating power of the frequency converter.

[0207] P i (t)=P base_i *R i (t);

[0208] Among them, P i (t) represents the real-time operating power of the propulsion device, P base_i This represents the baseline operating power of the power delivery device. Continuing with the previous calculation example, if the baseline operating power of the device is 10kW, the system calculates 10*1.3, resulting in a real-time operating power of 13kW. The system will then generate a corresponding 13kW power control command and send it to the physical execution layer.

[0209] In addition, the system has built-in physical constraints for equipment safety. When the calculated real-time operating power exceeds the rated maximum physical power of the motor, the system will forcibly cut off the excess and retain the highest rated power output.

[0210] In some alternative implementations, if the control hardware of the on-site power delivery device does not support continuous variable frequency control (CVC), the system can degrade the continuous adjustment logic into multi-level stepped control as an alternative. Specifically, the continuous adjustment factor is discretized into multiple fixed operating power levels, such as 30%, 60%, 100%, and 120%. Based on the instantaneous dynamic risk state range, the nearest or slightly higher fixed power level command is directly triggered, thus ensuring compatibility with traditional fixed-frequency physical equipment.

[0211] Compared to traditional fixed-power operation or binary switch control modes, the above servo control mechanism can significantly reduce equipment energy consumption during low-risk periods and smoothly accelerate reoxygenation response during high-risk periods. It should be understood that the specific energy-saving effect varies depending on the hydrological and meteorological conditions of the target lake and the area of ​​the high-risk zone.

[0212] Example 9 further details the physical layout of each propulsion device and the synergistic implementation of Venturi aeration. This example corresponds to the specific implementation scheme of the method at the physical execution layer and the engineering deployment layer. In one possible implementation, each propulsion device adopts a layered layout pattern in terms of spatial placement points, including the following steps:

[0213] Step 901: Deploy surface propulsion devices to improve the horizontal flow of water at the water surface of the target spatial locations.

[0214] The specific surface depth range is typically set to 0m to 1.0m below the water surface. The surface flow propulsion device can be a floating installation structure, with its impeller driven in the same direction as the prevailing wind direction or the reference flow field. The main function of this device is to create a horizontal artificial flow field, increasing the probability of disrupting water surface tension, accelerating atmospheric reoxygenation, and simultaneously using water flow shear force to inhibit the high-density aggregation of phytoplankton such as cyanobacteria on the water surface.

[0215] Step 902: Deploy a bottom-level flow propulsion device at the bottom of the water body where the target space is set up, to break the vertical static stratification of the water body and enhance the reoxygenation capacity of the bottom water body.

[0216] Bottom-layer flow propulsion devices are typically installed at a height of 0.5m to 1.5m above the lakebed sediment interface and can be fixed to the bottom of a guide rod using a submerged installation structure. Shallow lakes often form temperature stratification layers in summer, making it difficult for the bottom water to exchange oxygen with the surface oxygen-rich water. The impeller of the bottom-layer flow propulsion device is tilted upwards at a preset angle, which can range from 15° to 30°. By driving the bottom water upwards, it generates vertical mixing convection. This convection effectively disrupts the temperature stratification and density stratification, transporting dissolved oxygen from the surface to the bottom and inhibiting the metabolic activity of anaerobic microorganisms and the occurrence of sulfur-iron reduction reactions in the sediment.

[0217] Furthermore, the propulsion device is also equipped with an aeration device based on the Venturi principle.

[0218] To further increase the reoxygenation efficiency of the water body, the system integrates an aeration device in series at the outlet end of the bottom or surface flow propulsion device. The aeration device includes an inlet pipe, a contraction section, a throat, a diffuser section, and an air intake pipe connected to the throat, with the other end of the air intake pipe extending into the atmosphere above the water surface.

[0219] In some embodiments, adjusting the real-time operating power of each propulsion device according to the instantaneous dynamic risk status of the high-risk area of ​​lacustrine flooding further includes: while controlling the operation of the propulsion device, driving the aeration equipment to draw in air using the negative pressure generated by the high-speed water flow to form microbubble dissolved air water, so as to cooperate with the hydrodynamic circulation of the propulsion device to improve the overall reoxygenation efficiency of the water body in the high-risk area of ​​lacustrine flooding.

[0220] When the system drives the motor of the propulsion device according to the instantaneous dynamic risk state, the water flow is pressurized and injected at high speed into the contraction section of the aeration equipment. The negative pressure formed at the throat automatically draws air from the atmosphere into the pipe through the suction pipe, eliminating the need for an additional power-consuming air compressor. The air and high-speed water flow undergo intense shearing collisions in the diffusion section, breaking them down into microbubbles with diameters typically ranging from 0.1 mm to 2.0 mm. These microbubbles have a large specific surface area and a slow rising velocity, allowing oxygen to remain in the water for an extended period and fully dissolve. The dissolved air water, containing a large amount of dissolved oxygen and microbubbles, is then carried by the propulsion jet of the propulsion device into the high-risk lake area, achieving a synergistic effect of hydrodynamic flow generation and chemical oxygenation.

[0221] In some optional implementations, the intelligent management platform receives real-time equipment status data and operation logs from each node. For integrated propulsion and aeration equipment operating underwater, the system is equipped with a periodic anti-fouling maintenance control program. For example, after 72 hours of cumulative operation, the system can control the propulsion device's motor to run in reverse for 3 minutes continuously, using a water backwashing mechanism to remove algae and suspended debris attached to the impeller and venturi intake, thereby ensuring the hydraulic efficiency and air intake volume of the equipment during long-term operation.

[0222] According to one possible implementation of one aspect of this application, a method for preventing and controlling lacustrine flooding based on hydrodynamic regulation further includes the following steps:

[0223] S1 involves accurately defining high-incidence areas of lacustrine flooding by collecting and organizing historical flooding data, water quality monitoring data, hydrological and meteorological data, and sediment monitoring data for the target lake over the past 5-10 years. Historical flooding data includes the time, location, duration, and affected area of ​​the flooding. Water quality monitoring data includes key flooding indicators such as dissolved oxygen (DO), chlorophyll a, sulfides, and nitrogen and phosphorus content. Hydrological and meteorological data includes water temperature, wind speed, water level, and flow velocity. Sediment monitoring data includes sediment organic matter and sulfur and iron content. The above data is standardized and analyzed using spatial interpolation and frequency analysis methods, combined with the threshold conditions for flooding occurrence: dissolved oxygen ≤ 2.0 mg / L, water temperature ≥ 20℃, and stable weather conditions. This determines the specific range, boundary coordinates, and peak periods of high-incidence areas, and clarifies the topographic features, water depth, and weak points in hydrodynamics within these areas, providing precise targeting for subsequent hydrodynamic regulation.

[0224] S2 constructs a three-dimensional hydrodynamic-aquatic ecology coupled model of the lake. Based on the overall topographic data of the high-incidence lacustrine flooding area and the target lake defined in S1, a three-dimensional hydrodynamic-aquatic ecology coupled model of the lake is constructed. The model input parameters include lake topographic data, hydrological and meteorological data, water quality parameters, and sediment parameters. The hydrodynamic module simulates hydrodynamic characteristics such as water velocity, flow direction, water level, and mixing depth, while the aquatic ecology module simulates the migration and transformation processes of key lacustrine flooding indicators such as dissolved oxygen, chlorophyll a, and sulfides. The governing equations are closed by combining a two-equation turbulence model and solved using the mode splitting method to ensure that the model can accurately reflect the coupling relationship between hydrodynamic and aquatic ecology factors in the high-incidence lacustrine flooding area, providing scientific support for the determination of hydrodynamic regulation parameters.

[0225] S3 simulates the hydrodynamic demand in areas prone to lacustrine flooding. With lacustrine flooding prevention as the goal, the dissolved oxygen (DO) control threshold for these areas is set at 2.0 mg / L. This threshold effectively inhibits the reproduction of anaerobic microorganisms, blocks sulfur-iron reactions and the formation of blackening and odor-causing substances, preventing lacustrine flooding at its source. This threshold is input into the three-dimensional hydrodynamic-aquatic ecosystem coupling model constructed in S2. The model simulates the spatiotemporal distribution characteristics of dissolved oxygen in areas prone to lacustrine flooding under different hydrodynamic conditions (flow velocity, flow direction, mixing intensity). The optimal hydrodynamic parameters required to stably achieve a dissolved oxygen level >2.0 mg / L across the entire area are derived, including average water velocity, vertical mixing intensity, and horizontal diffusion rate. The model clarifies the differences in hydrodynamic regulation needs during different high-incidence periods (such as the hot and stable summer period and the autumn algal bloom decline period), forming targeted hydrodynamic regulation schemes to ensure the scientific validity and adaptability of the schemes.

[0226] S4. Precise Deployment and Operation of High-Fluidity Flow-Promoting Devices: Based on the optimal hydrodynamic parameters obtained from the S3 simulation, and combined with the topographic features, water depth, and weak points in the high-risk lacustrine flooding area, high-fluidity flow-promoting devices are precisely deployed within the high-risk area. These devices can generate strong regional water flows, achieving low-energy consumption and high-flow-rate reoxygenation. The devices are deployed in a multi-point distributed, layered pattern. Surface flow-promoting devices improve horizontal flowability, while bottom flow-promoting devices break down vertical water stratification and enhance the reoxygenation capacity of the bottom water. The deployment density is adjusted according to the gradient of hydrodynamic demand to ensure that after the flow-promoting devices are operational, the water in the high-risk lacustrine flooding area can reach the optimal hydrodynamic parameters determined in S3, achieving a stable dissolved oxygen level of >2.0 mg / L, thus preventing lacustrine flooding from the source. The devices can be installed in floating, submerged, or deep-water modes, requiring no special installation foundation, allowing for flexible movement, and minimizing impact on the water surface landscape.

[0227] S5 involves dynamic monitoring and hydrodynamic regulation optimization. Monitoring points are deployed in and around areas prone to lacustrine flooding. A combination of online monitoring and manual sampling is used to simultaneously monitor key lacustrine flooding indicators such as dissolved oxygen (DO), chlorophyll a, sulfides, flow velocity, and water temperature. Monitoring frequency is once every 2-4 hours, increasing to once every hour during peak periods. The effectiveness of lacustrine flooding control is judged based on the monitoring data: if dissolved oxygen remains consistently above 2.0 mg / L, and chlorophyll a and sulfide levels continue to decrease, with no black or odorous water, the control is considered effective. If the conditions are good, the operating power of the high-flux propulsion device can be appropriately reduced to decrease energy consumption. If dissolved oxygen is detected to drop to ≤2.0 mg / L, or chlorophyll a and sulfide content increases, indicating the beginnings of lake flooding, it indicates insufficient hydrodynamic control. In this case, the operating power of the propulsion device should be increased immediately, and the number of devices should be expanded to strengthen hydrodynamic control and ensure a rapid increase in dissolved oxygen content in the water body to block the occurrence and spread of lake flooding. At the same time, hydrodynamic control parameters should be adjusted according to seasonal changes and hydrological and meteorological conditions to achieve dynamic optimization and ensure the stability and economy of the control effect.

[0228] For example, taking a high-incidence area of ​​a lake as the verification object, the aforementioned three-dimensional hydrodynamic-aquatic ecology coupled model was used to simulate the spatiotemporal distribution of dissolved oxygen in the water body for 15 consecutive days during the summer high-temperature and stable period under the following three working conditions: Working condition 1, no propulsion device in operation; Working condition 2, propulsion device operating at a fixed rated power according to the traditional uniform deployment method; Working condition 3, propulsion device operating using the targeted deployment scheme and dynamic servo adjustment strategy determined by the method of this invention. The simulation results show that in working condition 3 of the method of this invention, the proportion of time in which the dissolved oxygen concentration in the bottom layer of the high-incidence area is stably maintained above the safety threshold is significantly higher than in working condition 2, and the cumulative power consumption of the propulsion device is lower than in working condition 2. It should be understood that specific performance indicators may vary depending on the topographic features, meteorological conditions, and equipment selection of the target lake.

[0229] This application systematically solves the problems of blind spatial deployment and time-delayed response in the prevention and control of lake flooding in large shallow lakes by deeply integrating nonlinear risk assessment, hydrodynamic model inversion and dynamic servo control.

[0230] By introducing second-order interaction terms to quantify the multi-factor synergistic outbreak effect and superimposing propagation corrections based on the baseline flow field, the true spatial pattern of risk is accurately depicted. Furthermore, by utilizing morphological indices and integer programming algorithms, abstract risk is transformed into a quantitative mapping of entity push flow density. This minimizes the global area-weighted residual risk under resource constraints, ensuring the targeted and accurate deployment of regulatory resources. This solves the problem of insufficient coverage of core high-incidence areas caused by the indiscriminate deployment of traditional equipment.

[0231] Employing a static-dynamic dual-layer architecture, the system incorporates real-time segmented mapping of multiple factors and dynamic weights calibrated based on historical pre-outbreak averages, enabling it to keenly capture very early risk evolution. Combined with segmented, continuous servo-controlled logic replacing traditional binary switch control, this eliminates the impact of frequent equipment start-ups and shutdowns on the power grid, achieving low-energy maintenance during low-risk periods and smooth, accelerated reoxygenation before high-risk outbreaks. This mechanism, combined with the construction of a stratified three-dimensional flow field on the bottom and surface and Venturi negative pressure microbubble aeration technology, fundamentally breaks down static stratification, blocks the formation of anaerobic environments, and improves the response speed and full-lifecycle reoxygenation efficiency of lake flood emergency response.

[0232] The preferred embodiments of the present invention have been described in detail above. However, the present invention is not limited to the specific details in the above embodiments. Within the scope of the technical concept of the present invention, various equivalent transformations can be made to the technical solutions of the present invention, and these equivalent transformations all fall within the protection scope of the present invention.

Claims

1. A method for preventing and controlling lacustrine flooding based on hydrodynamic regulation, characterized in that, include: Acquire multi-source environmental data for the target lake; Based on multi-source environmental data, a comprehensive risk assessment index for lacustrine flooding of the target lake is constructed, and the spatial risk gradient distribution of high-incidence lacustrine flooding areas in the target lake is defined accordingly. Using a three-dimensional hydrodynamic-aquatic ecology coupled model of the target lake, and with a preset safe threshold for dissolved oxygen concentration as a constraint, the target hydrodynamic control parameters required to bring the dissolved oxygen concentration to the safe threshold in the high-incidence area of ​​lake flooding are reverse-engineered. Based on the spatial risk gradient distribution and target hydrodynamic control parameters, the deployment density of propulsion devices in high-incidence lacustrine flooding areas and the spatial deployment locations of each propulsion device are determined. Real-time monitoring data of areas prone to lacustrine flooding is obtained, and the instantaneous dynamic risk status of these areas is determined based on the real-time monitoring data. The real-time operating power of each propulsion device is then adjusted accordingly.

2. The method according to claim 1, characterized in that, Multi-source environmental data includes water quality monitoring data, hydrological and meteorological data, and sediment monitoring data; Before constructing the comprehensive risk assessment index for lacustrine biota, the following data preprocessing steps are also included: Based on multi-source environmental data, we extract the basic risk factors of the target lake in each spatial grid cell. The basic risk factors are normalized and uniformly mapped to dimensionless risk scores.

3. The method according to claim 2, characterized in that, Construct comprehensive algal bloom risk assessment indicators for the target lake, including: For any spatial grid cell, based on its normalized basic risk factors, calculate the main effect risk contribution when each individual risk factor acts independently. Obtain pre-screened risk factor pairs with significant positive synergistic amplification effect, and calculate the synergistic interaction risk contribution of the risk factor pairs based on the product of the scores when the two basic risk factors co-occur. Based on preset interaction intensity adjustment parameters, the risk contribution of the main effect and the risk contribution of the synergistic interaction are nonlinearly weighted and fused to obtain the comprehensive risk assessment index for lake pansmography.

4. The method according to claim 3, characterized in that, After obtaining the comprehensive risk assessment index for lacustrine flooding, but before defining the spatial risk gradient distribution of high-incidence areas in the target lake, the process also includes correction for flow-induced risk propagation: Based on the three-dimensional hydrodynamic-aquatic ecology coupled model of the target lake, the baseline net flux of water flow between adjacent spatial grid cells in the target lake is extracted, and the upstream connected neighborhood of each spatial grid cell is determined accordingly. Based on the comprehensive risk assessment index of lake flooding in each spatial grid unit, and according to the preset connectivity influence coefficient, the risk increment of each unit in the upstream connected neighborhood after flux ratio weighting is iteratively accumulated to the local spatial grid unit until the iteration converges, and the final comprehensive risk assessment index of lake flooding after spatial propagation correction is obtained. Spatial clustering was performed based on the final comprehensive risk assessment index for lacustrine flooding to determine the continuous boundaries of high-incidence areas of lacustrine flooding and their spatial risk gradient distribution.

5. The method according to claim 1, characterized in that, The instantaneous dynamic risk status of areas prone to lacustrine flooding is determined based on real-time monitoring data, including: Obtain the preset absolute safety threshold and preset extreme danger threshold corresponding to each monitoring indicator in the real-time monitoring data; Using segmented risk conversion logic, the real-time physical measurement values ​​of each monitoring indicator are mapped to dimensionless instantaneous risk scores. Measurement values ​​that are at or above the preset absolute safety threshold are mapped to zero risk scores, while measurement values ​​that are at or below the preset extreme danger threshold are mapped to full risk scores. Obtain the dynamic calculation weights of each pre-configured monitoring indicator; The instantaneous risk scores of each monitoring indicator are weighted and summed with their corresponding dynamic calculation weights to obtain the instantaneous dynamic risk status that characterizes the sensitivity to multi-factor comprehensive triggering.

6. The method according to claim 2, characterized in that, Multi-source environmental data also includes remote sensing monitoring data; basic risk factors for the target lake are extracted in each spatial grid cell, including: From multi-source environmental data, the frequency factors of low oxygen in water, high temperature, stable weather, endogenous risk factors of sediment containing organic matter and sulfur and iron content, and high-density aggregation frequency factors of cyanobacteria detected by remote sensing images were extracted as basic risk factors.

7. The method according to claim 1, characterized in that, The propulsion devices are arranged in a layered pattern at their spatial locations, including: Surface propulsion devices are installed on the surface of the water body at the target spatial locations to improve the horizontal flow of the water; and, At the target spatial location, a bottom-level propulsion device is installed in the bottom layer of the water body to break up the vertical static stratification of the water body and enhance the reoxygenation capacity of the bottom water body.

8. The method according to claim 7, characterized in that, The propulsion device is also equipped with an aeration device based on the Venturi principle; Adjusting the real-time operating power of each propulsion device also includes: While controlling the operation of the propulsion device, the aeration equipment is driven to draw in air using the negative pressure generated by the high-speed water flow, forming microbubble dissolved air water, which, together with the hydrodynamic circulation of the propulsion device, improves the overall reoxygenation efficiency of the water body in the high-incidence area of ​​lacustrine flooding.

9. A system for preventing and controlling lacustrine flooding based on hydrodynamic regulation, characterized in that, include: At least one processor; as well as, A memory that is communicatively connected to at least one processor; wherein, The memory stores instructions that can be executed by a processor to implement a method for preventing and controlling lacustrine flooding based on hydrodynamic regulation as described in any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that, Computer-readable storage media include stored executable programs. The executable program, when running, controls the device containing the computer-readable storage medium to execute any one of the methods for preventing and treating lacustrine flooding based on hydrodynamic regulation as described in any one of claims 1 to 8.

Images