A riverway silt treatment method based on dynamic simulation and optimization decision

Abstract

The present application relates to the technical field of river dredging treatment, and relates to a river sludge treatment method based on dynamic simulation and optimized decision, obtains and processes historical and real-time multi-dimensional environmental data of a river to be treated and an influence area thereof, constructs a watershed process simulation system capable of simulating river water dynamics, sediment transport, pollutant migration and transformation and ecological response process, parameterizes definition of decision variables constituting a river treatment scheme, formal definition and setting of a multi-objective optimization problem, adopts a multi-objective optimization engine, embeds the watershed process simulation system as a core evaluation module into an optimization iteration loop, implements an optimized scheme, and continuously monitors environmental state changes during and after implementation, the present application can accurately predict long-term and short-term effects of different treatment schemes in terms of water dynamics, sediment, pollutant migration and transformation and ecological response, and solves the problem of sludge comprehensive treatment.

Description

Technical Field

[0001] This invention relates to the technical field of river dredging and treatment, and specifically to a method for river silt treatment based on dynamic simulation and optimization decision-making. Background Technology

[0002] Rivers are complex natural systems, and their health directly impacts regional flood control, drainage, navigation, ecological functions, and socio-economic development. Sedimentation, a crucial process in river evolution, has a dual impact on river function: moderate sediment deposition maintains riverbed stability and biodiversity, while excessive siltation leads to river channel shrinkage, reduced flood capacity, navigation obstruction, water quality deterioration, and even ecological degradation. Therefore, river dredging is a key measure to ensure river health and achieve sustainable utilization.

[0003] Traditional river dredging plans are typically based on survey data of a localized section of the river to be dredged, focusing on addressing the immediate siltation problem. The plan often relies on engineers' experience and simplified calculation methods. The core idea is to determine the dredging area and volume based on indicators such as silt thickness, and then arrange corresponding dredging equipment and operational schedules. While this method can alleviate localized siltation to some extent, it has the following insurmountable limitations:

[0004] For example, CN117552494B discloses a method for dredging rivers. By detecting various data of the river before dredging, the working parameters of the dredging device are determined. Based on the river data, suitable working parameters of the dredging device are determined. This method effectively ensures the dredging effect while avoiding damage to the riverbed during the dredging process. After the dredging is completed, the working parameters of the dredging device are adjusted to continuously clean the river, which can prevent the river from accumulating silt again and ensure that the river remains unobstructed.

[0005] However, rivers are dynamic, interconnected, and complex systems. The impacts of dredging operations on river flow, sediment transport, water quality, and ecology are not limited to the dredged area but can have a chain reaction upstream, downstream, and even across the entire basin through hydrodynamic and sediment transport processes. Traditional planning methods isolate dredging projects from the overall basin system, failing to fully consider the potential impacts of dredging operations on water and sediment processes in upstream and downstream sections, riverbed sediment-scour balance, water quality changes, and the ecological environment. There is a lack of accurate quantitative predictions of the dredging effects (such as the degree of sedimentation improvement and flood control capacity enhancement) and environmental impacts (such as increased downstream sedimentation, water quality deterioration, and ecological disturbance) at different times (e.g., short-term, medium-term, and long-term after dredging) and spatially (e.g., dredged section, upstream and downstream sections) after the implementation of different dredging schemes. This makes it difficult to scientifically evaluate the merits of different schemes during the scheme demonstration and comparison stages, resulting in a lack of scientific basis in the decision-making process and difficulty in accurately controlling the dredging effect and environmental risks. To address this issue, announcement number CN119250361A discloses a method, system, equipment, and storage medium for river dredging operations. By acquiring geographical, meteorological, hydrological, and silt data for the entire river section to be dredged, a basin-wide hydrodynamic-sediment mathematical model is constructed, providing a solid data foundation and model support for the scientific formulation of dredging operation schemes. Based on the basin-wide data, key parameters of the river section to be dredged are determined, and the proposed preliminary dredging operation scheme fully considers the characteristics of the river section itself, reflecting the concept of tailoring the scheme to local conditions. The preliminary scheme is input into the basin-wide model for dredging simulation, and the obtained dredging simulation data provides a quantitative basis for assessing the dredging effect and impact. Through multi-dimensional effect evaluation of the river section to be dredged, the effectiveness of the dredging scheme in improving river sedimentation, flood control and drainage, navigation conditions, water environment, and ecosystem can be accurately determined, achieving a scientific evaluation of the dredging effect.

[0006] The aforementioned techniques are used to evaluate the "preliminary dredging plan." This means that model simulation occurs after the preliminary plan is formed, and if multiple revisions and iterations are needed, the model simulation must be repeated each time to evaluate the revised plan. This iterative "plan generation - model evaluation - plan revision - model re-evaluation" model is inefficient, making it difficult to support rapid evaluation and comparison of a large number of potential plans, and also failing to provide direct, real-time guidance from model prediction results for the plan generation or optimization process. Summary of the Invention

[0007] Therefore, the purpose of this invention is to provide a river siltation treatment method based on dynamic simulation and optimization decision-making, which can accurately predict the long-term and short-term impacts of different treatment schemes on hydrodynamics, sediment, pollutant migration and transformation, and ecological response, and solve the problem of comprehensive siltation treatment.

[0008] To achieve the above objectives, the present invention provides the following technical solution:

[0009] A method for river siltation control based on dynamic simulation and optimization decision-making includes the following steps:

[0010] S1. Multi-source environmental data acquisition and feature analysis: Acquire and process historical and real-time multi-dimensional environmental data of the river to be treated and its affected areas. The data should include at least topography, hydrology and hydrodynamics, sediment characteristics, water quality and pollutant distribution, and ecosystem status data.

[0011] The data is fused and processed to extract key spatiotemporal features that characterize river siltation, water pollution, and ecological health.

[0012] S2. Construction and Validation of Watershed Process Simulation System: Construct a watershed-scale coupled watershed process simulation system capable of simulating river hydrodynamics, sediment transport, pollutant migration and transformation, and ecological response processes;

[0013] The key features extracted in step S1 and historical monitoring data are used to calibrate and verify the parameters of the watershed process simulation system to ensure its accuracy in simulating environmental processes.

[0014] S3. Parameterization and knowledge base support of river management scheme elements: Parameterization defines the decision variables that constitute the river management scheme. The decision variables include at least the dredging scope, dredging depth, dredging sequence, dredging equipment type and combination, dredged material treatment method, and pollutant degradation agent type, formula ratio, application location and dosage.

[0015] Construct a knowledge base that includes the relevant performance, cost, environmental impact, and applicable conditions of the above elements;

[0016] S4. Formal Definition of Multi-Objective Optimization Problem: Define multiple objective functions to evaluate the comprehensive benefits of river management schemes. The objective functions shall at least cover dredging efficiency, pollutant reduction rate, total project cost, environmental impact minimization, and ecological restoration effect. Define the technical, economic, environmental, and regulatory constraints that must be met during the operation.

[0017] S5. Multi-objective optimization engine driven by simulation feedback and scheme generation: A multi-objective optimization engine is adopted, and the watershed process simulation system is embedded into the optimization iteration loop as its core evaluation module;

[0018] The multi-objective optimization engine, based on the defined decision variables and the defined multi-objective optimization problem, calls the watershed process simulation system in real time during its search process to simulate the environmental state evolution and expected results of multi-objective evaluation indicators after the implementation of different candidate governance schemes. It also dynamically adjusts its search strategy based on the prediction feedback of the simulation system to generate a set of river governance schemes that meet the constraints and make the multi-objective function optimal.

[0019] S6. Implementation, Dynamic Monitoring and Adaptive Adjustment: Implement the preferred scheme generated in step S5, and continuously monitor changes in environmental conditions during and after implementation; compare and analyze the monitoring results with the prediction results of the watershed process simulation system; if significant deviations or new environmental disturbances occur, trigger the updating of the parameters or structure of the watershed process simulation system, and selectively rerun the optimization process of step S5 to generate an adjustment scheme adapted to the new environmental conditions.

[0020] The present invention is further configured such that the watershed process simulation system integrates at least the following sub-models:

[0021] Hydrodynamics Module: Based on the principles of fluid mass conservation and momentum conservation, this module uses two-dimensional or three-dimensional numerical methods to solve the governing equations to simulate the hydrodynamic processes of water level, velocity field, and flow distribution in rivers and related water areas driven by water flow, tides (if applicable), wind fields, etc., under specific river topographic boundaries and riverbed roughness conditions.

[0022] Sediment transport module: This module is bidirectionally coupled with the hydrodynamic module. Based on the water flow conditions and the physical properties of sediment (including particle size distribution, density, settling velocity, etc.), it simulates the convection, diffusion, and settling processes of suspended sediment and the initiation and transport processes of bedload sediment, and considers the composition of riverbed sediment and the influence of critical shear stress.

[0023] Riverbed Evolution Module: This module is based on the principle of sediment continuity and dynamically updates the riverbed topography and elevation based on the amount of sediment erosion and deposition per unit time calculated by the sediment transport module.

[0024] The present invention is further configured such that: in the sediment transport module, the calculation of the erosion rate and deposition rate of riverbed sediment is further limited to:

[0025] The erosion rate is calculated based on the comparison between the actual shear stress of the water flow acting on the riverbed and the critical erosion shear stress of the sediment. When the actual shear stress is greater than the critical erosion shear stress, sediment erosion occurs. The erosion rate is related to the difference or ratio between the two and the properties of the sediment.

[0026] The sedimentation rate is calculated based on the sedimentation characteristics of the sediment and the sediment concentration in the near-bed flow. It can also take into account the comparison between the actual shear stress of the flow and the critical sedimentation shear stress of the sediment. When the actual shear stress is less than the critical sedimentation shear stress, it is conducive to sediment deposition.

[0027] The present invention is further configured such that: the watershed process simulation system further integrates a pollutant migration and transformation module, which is coupled with a hydrodynamic module and a sediment transport module to simulate the following processes of one or more target pollutants in the study water area:

[0028] The distribution of pollutants between the dissolved phase and the particulate phase, including adsorption and desorption processes;

[0029] Convective transport of dissolved and adsorbed pollutants with water flow and diffuse transport caused by concentration gradients;

[0030] Adsorbed pollutants are deposited on the riverbed as suspended sediment settles, and pollutants already deposited on the riverbed are re-entered into the water as bottom sediment is resuspended.

[0031] The biochemical degradation or transformation process of pollutants themselves, the rate of which is affected by factors such as the nature of the pollutants, environmental conditions, and microbial populations;

[0032] The present invention is further configured such that: the parameterized definition of the elements of the river management scheme includes:

[0033] Spatial parameters for dredging: These include a sequence of geographic coordinate points or a set of grid cells that define the boundaries of the dredging area, as well as the target post-dredging riverbed elevation or planned dredging thickness set within each area or cell;

[0034] Dredging time parameters: including the planned start date of each dredging area, the expected number of days or hours of continuous operation, and the logical relationship between the operation sequence of different areas;

[0035] Dredging process and equipment parameters: The model index, required quantity, and key operational parameters of a specific dredging vessel selected from the knowledge base;

[0036] Dredged material disposal and resource utilization parameters: selected dredged material transfer or final disposal site number, transport distance, acceptable dredged material characteristics, and selected resource utilization process type and its corresponding processing capacity and cost coefficient;

[0037] Pollutant degradation agent application parameters:

[0038] Degradation agent selection and compatibility: Identifiers for specific biological or chemical agents targeting the pollutant; if it is a compound, it includes the precise mass or volume percentage of each component.

[0039] Spatial application strategy: The application area of ​​the degradative agent is defined in the form of a two-dimensional grid, and the spatial distribution function of the planned application dose or total application amount of each grid cell;

[0040] Application time strategy: the start time, duration and frequency of application of the degradation agent. The application frequency includes single shock dose, multiple pulse dose or continuous low dose.

[0041] The present invention is further configured such that: the objective function of the multi-objective optimization problem has a specific mathematical expression or its constituent terms that include at least:

[0042] Maximize the completion rate of dredging work: ;

[0043] in It is a plan The predicted actual dredging volume for the i-th dredging area. It is the target dredging volume. This represents the total number of dredged areas. It is the weighting coefficient.

[0044] Maximizing the benefits of pollutant reduction: ,in, It is the initial total amount of the j-th target pollutant. It is a plan The final total amount of the pollutant predicted after implementation, It is the number of target pollutant types. These are the weighting or benefit conversion factors for the reduction of different pollutants.

[0045] Minimize total economic cost: The costs are fixed costs, dredging operation costs, transportation costs, disposal costs, degradation agent costs, and monitoring costs, minus the potential revenue from the resource utilization of dredged materials.

[0046] Minimize environmental risks: ,

[0047] in It is a plan The predicted k-th negative environmental impact indicator includes, specifically, the peak concentration of suspended solids at a specific downstream section, the short-term water quality deterioration index caused by dredging disturbance, and the risk of exceeding the standard concentration of degradation agent residues or by-products. It refers to the number of environmental impact indicators. It's the weight.

[0048] Maximizing the restoration of ecological functions:

[0049] ,

[0050] in It is a comprehensive ecological evaluation function whose inputs are multiple predicted ecological indicators, such as improved water transparency, increased dissolved oxygen levels, potential for recovery of benthic biodiversity index, and increased coverage of specific indicator aquatic plants.

[0051] The present invention is further configured such that the constraints of the multi-objective optimization problem, in their specific mathematical expression or logical judgment, include at least the following:

[0052] Budget ceiling: Total economic cost ;

[0053] Overall project timeline constraints: The operation time and equipment dispatch time in each area were taken into account.

[0054] Channel function assurance: For any point (x, y) within the channel area, the water depth after dredging ;

[0055] Environmental quality standard compliance: For all pre-set environmental control points m, the predicted concentrations of all key water quality parameters q during the operation and recovery periods. ;

[0056] Ecological red line area protection: areas where dredging or degradation agents are applied. and ecological protection red line areas There is an intersection Then it requires Or the intensity of operations / disturbance level within this intersection area. ,

[0057] Dredging disposal site capacity: Scheme Total volume of dredged material produced ,

[0058] Degrader safety threshold: The predicted concentration of the parent degrader or its known major degradation intermediates at any point in the environment. ,

[0059] Riverbank or levee stability: The minimum safety factor for adjacent riverbanks or levees after dredging. .

[0060] The present invention is further configured such that: the multi-objective optimization engine is based on a heuristic algorithm of population evolution. This algorithm maintains a population of candidate river management schemes and iteratively improves the quality of the schemes in the population by repeatedly applying evolutionary operators such as selection, crossover, and mutation. Specifically, the implementation method is as follows:

[0061] Initialization: An initial population containing multiple different river management schemes is generated randomly or according to empirical rules. Each scheme is defined by a vector of decision variables. express;

[0062] Fitness assessment: For each individual in the population, the watershed process simulation system is invoked to predict the environmental evolution process after the implementation of the scheme, and multiple objective function values ​​are calculated based on this prediction. ;

[0063] Selection: Based on the multi-objective function value of an individual, non-dominated sorting and crowding calculation are used to select superior individuals from the current population to enter the next generation or to participate in reproduction as parents.

[0064] Crossover: Select parent pairs of individuals and their decision variable vectors. Crossover operations involve exchanging or combining some or all of the elements to generate new offspring individuals. The design of crossover operations should consider the characteristics of the decision variables.

[0065] Variation: The vector of decision variables for offspring individuals Small random perturbations or changes are made to certain elements in order to maintain population diversity and explore new regions of the solution space;

[0066] Elite strategy: Store the non-dominated solutions generated in each generation in an external archive and ensure that these solutions can be passed on to the next generation to prevent the loss of excellent solutions;

[0067] Termination conditions: When the preset maximum number of iterations, computation time limit, and the quality of solutions in the population no longer improve, the termination condition is reached.

[0068] The present invention is further configured such that: the search strategy of the multi-objective optimization engine based on simulated feedback dynamically adjusts at least includes:

[0069] Adaptive operator probability adjustment: Records the frequency or utility of the algorithm in the past few generations in successfully producing better solutions through crossover and mutation operators. If a certain operator or its specific parameter settings have performed better recently, the probability of its selection is dynamically increased or its parameter range is adjusted; otherwise, it is decreased.

[0070] Search bias adjustment based on solution density and distribution: Real-time analysis of the distribution density of current Pareto front solutions in the target space. If solutions are too dense in certain regions, the priority of exploring these regions is reduced when selecting a parent or generating a new individual.

[0071] The present invention is further configured such that updating the parameters or structure of the watershed process simulation system includes:

[0072] Online correction of specific physical parameters: Using Kalman filtering or variational assimilation methods, the acquired real-time monitoring data, including water level, flow velocity, measured suspended sediment concentration, and key pollutant concentration at a specific cross-section, are compared with the corresponding outputs of the simulation system. By minimizing the deviation between the two, key physical parameters in the simulation system are retrieved and updated online, such as the riverbed roughness coefficient and the critical initiation shear stress of sediment. Comprehensive degradation rate coefficient of pollutants or adsorption partition coefficient ,

[0073] Model structure diagnosis and adjustment based on error propagation analysis:

[0074] By analyzing the temporal and spatial distribution characteristics of the prediction error of the simulation system, and combining with sensitivity analysis, we can identify which model assumptions, simplified control equations, or missing subprocesses may be the main sources of error.

[0075] If the diagnosis indicates that the existing model structure cannot adequately describe the observed phenomena, the model structure is adjusted, including upgrading from a two-dimensional model to a three-dimensional model, or introducing new reaction kinetic equations or coupling new ecological modules into the pollutant migration and transformation module.

[0076] Compared with the shortcomings of the prior art, the beneficial effects of the present invention are as follows:

[0077] This method can accurately predict the long-term and short-term impacts of different treatment schemes on multiple aspects, including hydrodynamics, sediment, pollutant migration and transformation, and ecological response. Furthermore, by employing a multi-objective optimization algorithm, it can systematically weigh multiple objectives such as dredging effectiveness, pollutant reduction, economic costs, environmental risks, and ecological restoration, thereby generating river silt treatment schemes that achieve better overall benefits.

[0078] By embedding a complex watershed process simulation system as the core evaluation module into a multi-objective optimization loop, deep coupling between simulation and optimization is achieved. The optimization algorithm can call the simulation system in real time to evaluate the effectiveness of candidate solutions and dynamically adjust the search strategy based on feedback, thereby efficiently exploring a large number of potential solutions and overcoming the inefficiency of the traditional iterative mode of "solution generation-model evaluation-solution correction-model re-evaluation". Attached Figure Description

[0079] Figure 1 This is a schematic diagram of the process of the present invention;

[0080] Figure 2 This is a block diagram of the watershed process simulation system of the present invention. Detailed Implementation

[0081] Reference Figures 1 to 2 The embodiments of the present invention will be further described below.

[0082] S1. Multi-source environmental data acquisition and feature analysis: Historical and real-time multi-dimensional environmental data will be collected. This data needs to cover all aspects of the river to be treated and its affected areas in order to comprehensively reflect the state of the river system.

[0083] Topographic and geomorphological data: including high-precision DEM (Digital Elevation Model), river channel cross-section measurement data, and riverbed sediment type distribution data. These are acquired through technologies such as satellite remote sensing, UAV mapping, and sonar detection. The data should be detailed, including river channel morphology, slope, deep channel location, floodplain distribution, and bank slope characteristics.

[0084] Hydrological and hydrodynamic data: including historical and real-time data on water level, flow rate, velocity, and tidal level (if applicable). This data can be obtained by setting up hydrological stations, current meters, and tide gauges, or by utilizing historical monitoring records. The data frequency should be high enough to capture dynamic changes in water flow.

[0085] Sediment characteristics data include particle size distribution (e.g., proportions of clay, silt, fine sand, medium sand, and coarse sand), density, water content, organic matter content, critical initiation shear stress, critical deposition shear stress, and settling velocity of the sediment. Suspended sediment concentration and particle size distribution data also need to be collected. These data are obtained through sediment sampling, suspended sediment sampling, and laboratory experimental analysis.

[0086] Water quality and pollutant distribution data include the concentrations, pH, water temperature, conductivity, and transparency of dissolved oxygen (DO), chemical oxygen demand (COD), biochemical oxygen demand (BOD), ammonia nitrogen (NH4-N), total phosphorus (TP), total nitrogen (TN), heavy metals (such as Cu, Zn, Pb, Cd, Hg, etc.), and other potential pollutants (such as persistent organic pollutants (POPs)). Data is obtained through fixed-point monitoring, mobile monitoring, and automatic monitoring stations. Simultaneously, it is necessary to investigate the type, location, discharge volume, and discharge patterns of pollution sources.

[0087] Ecosystem status data includes benthic biodiversity indices, fish diversity and abundance, aquatic vegetation types and coverage, plankton community structure, and vegetation cover in river corridors. Data were obtained through ecological sampling surveys, biological monitoring, and remote sensing image analysis.

[0088] Data fusion processing and feature extraction:

[0089] Fusion processing: Data from different sources, in different formats, and with different spatiotemporal resolutions are processed through unified coordinate system transformation, interpolation, and correction to establish an integrated spatial database or geographic information system (GIS).

[0090] Feature extraction: Based on the fused data, key spatiotemporal features characterizing river siltation, water pollution, and ecological health status are calculated.

[0091] Siltation characteristics: historical riverbed elevation change rate (reflecting long-term siltation trend), existing siltation thickness distribution map, siltation volume in a specific area, spatial variability of sediment particle size distribution, and distribution of bottom sediment organic matter content (reflecting the condition of polluted silt).

[0092] Pollution characteristics: distribution and concentration of key pollutants (such as heavy metals and organic pollutants) in sediment and water bodies, spatial range of highly polluted areas, pollution load estimation, and spatiotemporal changes of water quality indices (such as the comprehensive pollution index WPI and the eutrophication index TSI).

[0093] Ecological health characteristics: biodiversity index of a specific area, distribution of indicator species, aquatic vegetation coverage and health status, and dissolved oxygen saturation distribution (reflecting the self-purification capacity of water bodies).

[0094] S2. Construction and Validation of the Watershed Process Simulation System: Build a watershed-scale coupled model capable of simulating complex river processes. This system needs to integrate multiple sub-modules, including hydrodynamics, sediment transport, riverbed evolution, pollutant migration and transformation, and ecological response, and achieve bidirectional coupling between these modules.

[0095] Composition of a watershed process simulation system:

[0096] Hydrodynamics module: Based on two-dimensional or three-dimensional unsteady flow governing equations (mass conservation equations and momentum conservation equations, i.e., Navier-Stokes equations or their simplified forms), numerical methods such as finite difference, finite volume, or finite element methods are used to solve them.

[0097] Mass conservation equation (two-dimensional):

[0098]

[0099] Momentum conservation equation (two-dimensional):

[0100] X direction:

[0101]

[0102] Y direction:

[0103]

[0104] Where H is the water depth, u and v are the components of the average flow velocity in the x and y directions, respectively, Z is the riverbed elevation, and g is the acceleration due to gravity. It is the density of water. It is pore water pressure. , , , It is the Reynolds stress tensor. , These are the components of the shear stress at the bottom of the riverbed in the x and y directions. It is the Coriolis parameter. It is a source and sink item. , Other external forces (such as wind stress).

[0105] The model needs to take into account river topography, riverbed roughness (Manning coefficient or Chezy coefficient), and boundary conditions (inflow rate, outflow level, tidal process, etc.).

[0106] Sediment transport module:

[0107] It is coupled with a hydrodynamic module to utilize information such as flow velocity, water depth, and riverbed shear stress obtained from simulations.

[0108] Suspended mass transport equation (two-dimensional):

[0109]

[0110] in, It is the concentration of suspended sediment. , It is the dispersion coefficient of suspended sediment. It is the source and sink of suspended sediment, mainly including the erosion and deposition processes of the riverbed.

[0111] Calculation of erosion and deposition rates of riverbed sediment:

[0112] Erosion rate: when the actual shear stress Greater than the critical erosion shear stress At that time, erosion occurs. Formula: , where M is the erosion coefficient, which is related to the properties of the sediment.

[0113] Deposition rate: when the actual shear stress Less than the critical deposition shear stress At this time, deposition is more likely. Formula: ,in It is the sediment settling rate. It refers to the concentration of suspended sediment near the bottom of the bed.

[0114] Bedload transport: The bedload transport rate is calculated based on flow conditions (such as Chezy stress or bed shear stress). Commonly used formulas include the Meyer-Peter & Müller formula and the Ashida & Michiue formula.

[0115] Riverbed evolution module:

[0116] Based on the principle of sediment continuity, the change in riverbed elevation is calculated according to the net sediment transport rate of suspended and bedload.

[0117] Equation for the rate of change of bed surface:

[0118] in, It is the dry density of the bed sand. , It is the component of bedload transport rate in the x and y directions. It is the net deposition rate of suspended matter per unit area (deposition minus erosion).

[0119] Pollutant migration and transformation module: coupled with hydrodynamics and sediment transport module.

[0120] Pollutant transport equation (dissolved phase, two-dimensional):

[0121] ;

[0122] Pollutant transport equation (adsorbed phase, transport with suspended sediment): ;

[0123] in, It is the concentration of pollutants in the dissolved phase. It is the concentration of pollutants in the adsorbed phase (the mass of pollutants adsorbed per unit mass of suspended sand). It is the rate of adsorption of the dissolved phase onto the particulate phase. , These represent the reaction or degradation rates of pollutants in the dissolved and adsorbed phases, respectively. , It is a source of pollutants.

[0124] Adsorption / desorption processes (such as linear adsorption, Freundlich adsorption, Langmuir adsorption), biochemical degradation kinetics (such as first-order reaction, Monod kinetics), volatilization, precipitation, and other processes need to be considered.

[0125] Ecological Response Module (optional, depending on specific needs): Simulates the response of ecological elements such as dissolved oxygen, algae, benthic organisms, and aquatic vegetation in water bodies to changes in the aquatic environment (hydrodynamics, sediment, pollutants, light, temperature, etc.). The dissolved oxygen module needs to consider reaeration, BOD degradation, sediment oxygen consumption, plant photosynthesis, and respiration.

[0126] Construction: Select Delft3D, EFDC, and WASP. Based on the collected topographic, hydrological, sediment, and pollution source data, construct the model calculation grid, define boundary conditions, and input parameters.

[0127] Calibration and verification: The simulation system is calibrated and verified using historical monitoring data.

[0128] Calibration involves adjusting sensitive parameters in the model (such as riverbed roughness, critical shear stress of sediment, pollutant reaction rate coefficient, and distribution coefficient) to achieve the best fit between the simulation results and historical monitoring data (such as water level, flow rate, suspended sediment concentration, pollutant concentration, and riverbed elevation changes). Commonly used calibration methods include trial and error and automatic calibration algorithms (such as SCE-UA and DREAM).

[0129] Validation: Using a separate set of independent monitoring data, without changing the calibrated parameter values, run the model and compare the simulation results with the validation data. Commonly used evaluation metrics include root mean square error (RMSE), Nash-Satcliffe efficiency coefficient (NSE), and coefficient of determination (R²). Only through rigorous calibration and validation can the predictive ability of the model be ensured.

[0130] S3. Parameterization and knowledge base support of river management scheme elements: Deconstruct complex river management schemes into a set of quantifiable decision variables and establish a knowledge base to support the selection of these decision variables.

[0131] Parameterization of decision variables:

[0132] Dredging range: This can be parameterized as a series of polygonal regions (defined by vertex coordinates) or a set of grid cells. Each region or cell can have a Boolean variable (whether to dredge) and a target post-dredging riverbed elevation or planned dredging thickness. For example, define region 1 (Polygon1) to be dredged to an elevation of -5m, and region 2 (Polygon2) to a dredging thickness of 1.5m.

[0133] Dredging depth: For the selected dredging area, define the target post-dredging riverbed elevation or the planned thickness of the mud layer to be removed.

[0134] Dredging Sequence: Define the start date, duration, and priority of dredging operations for different areas. For example, Area 1 will begin on July 1, 2024, and last for 30 days; Area 2 will begin after Area 1 is completed and last for 45 days.

[0135] Dredging Equipment Types and Combinations: Based on the properties of the silt to be dredged (e.g., viscous, loose), water depth, and environmental requirements (e.g., low disturbance), select suitable dredging equipment (e.g., grab dredgers, cutter suction dredgers, jet dredgers, underwater robotic dredging equipment, etc.) and their combinations from the knowledge base. Parameters can include equipment model index, quantity, and key operating parameters (e.g., cutter head speed, pumping flow rate).

[0136] Dredged material treatment method: Select the transportation method of the dredged material (pipeline transport, ship transport, truck transport), the transit or final disposal site (land-based storage yard, underwater disposal area), and whether to carry out resource utilization (such as brick making, composting, building materials). Parameters can include site number, transportation route, and resource utilization process type.

[0137] Types, formulation ratios, application locations, and dosages of pollutant degrading agents:

[0138] Degrading agent selection and compatibility: Select a degrading agent product identifier from the knowledge base that targets a specific pollutant (such as a heavy metal passivator or an organic matter degrading microbial agent). If it is a compound formulation, include the mass or volume percentage of each component (such as strain A, strain B, and nutrient C).

[0139] Spatial application strategy: Define the spatial region for degradation agent application (e.g., a specific contaminated sediment area, a contaminated water body area), and parameterize the planned application dose (degradation agent mass / volume per unit area / volume) for each region or grid cell. This can be defined as discrete points, polygonal regions, or a continuous function based on the pollution concentration distribution.

[0140] Application time strategy: Define the start time, duration, and frequency of application of the degradation agent. For example, the first application can be given one week before dredging and last for 3 days; or after dredging, it can be given once every two weeks for 3 months.

[0141] Knowledge base construction: Establish a database containing information related to the above decision variables.

[0142] Information on dredging equipment: performance parameters (maximum dredging depth, productivity), applicable conditions (sludge type, water depth), cost (rental / purchase costs, operating energy consumption, maintenance costs), environmental impact (degree of disturbance, noise), and historical application cases for different models of equipment.

[0143] Dredged material treatment information: capacity, receiving standards, processing capacity, processing costs, and environmental permit information for different treatment sites. Applicable sediment characteristics, processing costs, product types, and environmental benefits for different resource utilization processes.

[0144] Information on pollutant degradation agents: Mechanism of action of different types of degradation agents, applicable pollutant types, recommended formulations, optimal environmental conditions (temperature, pH), application methods, effective duration, cost, potential environmental risks (such as impacts on non-target organisms), and successful application cases.

[0145] S4. Formal definition of multi-objective optimization problem: Transform the evaluation objective of river management scheme into a mathematically computable objective function, and clarify the constraints that the scheme must satisfy.

[0146] Objective function: Set multiple objectives that may conflict with each other, and use them to evaluate the merits of different solutions.

[0147] Maximizing the completion rate of dredging work: This aims to maximize the ratio or difference between the actual completed dredging volume and the planned target dredging volume.

[0148] Mathematical expression: ;

[0149] in, It is a plan The actual dredging volume predicted in the i-th dredging area is obtained by calling the watershed process simulation system. This is the target dredging volume set in the regional plan. It is the total number of all preset dredging areas. These are weighting coefficients that reflect the importance of different dredging areas; for example, the weight of a waterway area may be higher than that of a non-waterway area. This objective function encourages the optimization engine to find a solution that maximizes the actual dredging volume and makes it as close as possible to the target volume.

[0150] Substitution calculation steps: For example, suppose there are two dredging areas, and the target dredging volumes are respectively and Weight =1 / 2. A certain plan. Through prediction models

[0151] The actual dredged volume is and .

[0152] Calculate the completion rate of the i-th region: ; ;

[0153] Calculate the objective function value: ;

[0154] Maximizing this value indicates that the dredging completion rate of the plan should be as high as possible.

[0155] Maximizing the benefits of pollutant reduction: This aims to maximize the amount or rate of reduction of the target pollutant.

[0156] Mathematical expression: in, It is the first The total mass of the target pollutants (such as organic matter, heavy metals, nitrogen, phosphorus, etc.) in the entire affected water area (water body and sediment) before the implementation of the plan.

[0157] Detailed explanation: This value was determined through data collection and analysis. It is a plan After implementation, the total mass of the pollutant at a specific point in time (e.g., after dredging is completed and a certain recovery period has elapsed) is obtained through simulation and prediction. It refers to the number of target pollutant types. These are the weights or benefit conversion factors for reducing different pollutants. For example, reducing a unit mass of toxic heavy metals may receive a higher benefit weight than reducing a unit mass of organic matter. This objective function prompts optimization algorithms to find solutions that can significantly reduce the total amount of pollutants in the river system.

[0158] Substitution calculation steps: Assume that the focus is on two pollutants: organic matter (expressed as COD) and heavy metal Cd. =2, initial total COD mass Initial total Cd mass A certain solution Through simulation and prediction, the final total COD mass is 400 kg, and the final total Cd mass is 20 kg. Weights are then set. , ,

[0159] Calculate the pollutant reduction amounts: COD reduction amount 1000 - 400 = 600kg; Cd reduction amount 50 - 20 = 30kg;

[0160] Calculate the objective function value: *600+ *30=900 (yuan benefit), maximize this value.

[0161] Minimizing total economic cost: This aims to find the most cost-efficient governance solution. Mathematical expression: ,

[0162] These are fixed costs that are not closely related to the specific content of the plan, such as preliminary research and report preparation. According to the plan The dredging cost is calculated using a knowledge base and engineering experience, taking into account the dredging scope, depth, equipment type, operation time, and sediment properties. The transportation cost is calculated based on the volume of dredged material, the transportation distance, and the mode of transportation (ship, truck, etc.). The disposal cost is calculated based on the volume of dredged material, the disposal method (stockpiling, landfilling, solidification, etc.), and the standard cost of the disposal site. The cost of the degradation agent is calculated based on the type, dosage, and frequency of application of the degradation agent in the plan, as well as the unit price in the knowledge base. The monitoring cost is calculated based on the monitoring items, frequency, and locations required by the plan. The objective function is the potential revenue generated from the resource utilization of dredged materials (such as brick making or soil for landscaping) in the proposed solution. This objective function encourages the optimization engine to find the solution with the minimum total investment.

[0163] Substitution calculation steps: Assume a certain solution The costs and benefits involved are as follows:

[0164] Fixed costs =500,000 yuan;

[0165] Dredging operation costs =2 million yuan (for example, dredging volume) = (Unit cost: 100 yuan / m³);

[0166] Transportation costs =800,000 yuan;

[0167] disposal costs =1.5 million yuan;

[0168] Cost of degradation agent =300,000 yuan;

[0169] Monitoring costs =100,000 yuan;

[0170] Benefits of dredged material resource utilization =200,000 yuan.

[0171] Calculate the objective function value: 50 + 200 + 80 + 150 + 30 + 10 - 20 = 5 million yuan; minimize this value.

[0172] Minimize environmental risks: This aims to reduce the negative environmental impacts during and after the implementation of the plan.

[0173] Specifically:

[0174] in It is a plan The k-th negative environmental impact indicator is predicted. These indicators can be specific values, such as the maximum peak value of suspended solids concentration at downstream drinking water source sections during dredging; or they can be comprehensive indices calculated from multiple indicators, such as the short-term water quality deterioration index (considering DO, turbidity, ammonia nitrogen, etc.). Other indicators may include the area of ​​the region where the residual concentration of degradation agents exceeds the standard, the area of ​​disturbance to the habitat of sensitive aquatic organisms, and the degree of impact of construction noise on surrounding residents.

[0175] Specifically, this includes peak concentrations of suspended solids at specific downstream sections, short-term water quality deterioration indices caused by dredging disturbances, and the risk of exceeding standards for residual degradation agents or byproduct concentrations. It refers to the number of environmental impact indicators. It is a weight that reflects the importance or severity of different environmental risks. For example, the water quality risk weight of drinking water sources should be very high.

[0176] Substitution-based calculation steps: Assume that we are interested in the peak concentration of downstream suspended solids and the recovery time of benthic organisms in the dredged area. =2, a certain scheme Based on simulation predictions, the peak downstream suspended solids concentration is 150 mg / L, and the benthic organism recovery time is 180 days. Weights were then set. =0.1 (yuan / (mg / L)) (For example, each additional day of recovery time results in a loss of 5 yuan), calculate the objective function value: a loss of 915 yuan, and minimize this value.

[0177] Maximizing ecological function restoration: This aims to promote the health and recovery of river ecosystems.

[0178] formula: ,

[0179] This is a comprehensive ecological evaluation function that takes multiple ecological indicators obtained through simulation and prediction as input and outputs a comprehensive evaluation score. These ecological indicators can be the average improvement in water transparency, the increase in average dissolved oxygen saturation, the predicted value of benthic biodiversity index at a specific time point (e.g., percentage relative to a reference state), the predicted coverage area of ​​specific indicator aquatic plants, fish habitat suitability index, etc. The comprehensive ecological evaluation function can be a weighted sum of these indicators, a multi-level fuzzy comprehensive evaluation, or a higher-level assessment based on the output of the ecological model. This objective function guides the optimization algorithm to find the most beneficial solution for ecosystem restoration.

[0180] Substitution calculation steps: Assume a comprehensive ecological evaluation function The degree of improvement in water transparency (using the predicted depth of improvement) (represented by) and benthic biodiversity index (using predicted relative values) The weighted sum of (indicated by) =2,

[0181] ,

[0182] A certain solution Through simulation predictions, water transparency is expected to improve. = 1.5m, benthic biodiversity index reaches 0.6 of the reference state, weights are set. =0.4, =0.6;

[0183] The objective function value is calculated as follows: 0.4*1.5+0.6*0.6=0.96. Maximizing this value indicates that the higher the degree of ecological function restoration, the better.

[0184] Constraints: Define the various restrictions that a solution must meet. These constraints limit the search space of solutions to a practically feasible and compliant range.

[0185] Budget ceiling: Total economic cost The total cost of the plan cannot exceed the preset budget limit.

[0186] Overall project timeline constraints: The total execution time of the plan (including dredging operation time, degradation agent application time, and equipment dispatch, entry and exit time, etc.) must not exceed the preset total project duration requirement. It is the operating time of the i-th region, calculated through simulation or based on equipment performance.

[0187] Channel function assurance: For any point (x, y) within the channel area, the water depth after dredging The water depth calculated from the simulated predicted riverbed elevation after dredging must meet the minimum navigation channel depth requirements at that point.

[0188] Environmental quality standard compliance: For all pre-set environmental control points m, the predicted concentrations of all key water quality parameters q during the operation and recovery periods. The concentrations of water quality parameters at each time step of the control points obtained through simulation prediction must meet the corresponding environmental quality standards.

[0189] Ecological red line area protection: areas where dredging or degradation agents are applied. and ecological protection red line areas There is an intersection Then it is required Or the intensity of operations / disturbance level within this intersection area. This means that operations in sensitive ecological areas must either be completely avoided, or the intensity of operations must be strictly limited to permissible levels.

[0190] Dredging disposal site capacity: Scheme Total volume of dredged material produced The predicted total volume of dredged sludge must not exceed the capacity of the designated disposal site.

[0191] Degrader safety threshold: The predicted concentration of the parent degrader or its known major degradation intermediates at any point in the environment. The concentration of the degradation agent and its key metabolites in water or sediment, as predicted by simulation, must not exceed the set safety threshold to ensure that it is harmless to the aquatic ecosystem and human health.

[0192] Riverbank or levee stability: After dredging, the stability of adjacent riverbanks or levees is assessed by combining simulated and predicted riverbed morphology changes with soil mechanics calculations, and a minimum safety factor is determined. To ensure that dredging does not cause bank collapse or affect the safety of flood control facilities.

[0193] Other constraints include avoiding specific fishing ban periods and migratory bird seasons, and ensuring that construction noise meets standards.

[0194] S5. Multi-objective optimization engine driven by simulation feedback and scheme generation: The constructed complex watershed process simulation system serves as an "online evaluator," directly embedded into the iterative loop of the multi-objective optimization algorithm. The optimization algorithm is no longer independent of the simulation; instead, each generated candidate scheme is immediately evaluated through the simulation system, and the search direction is dynamically adjusted based on the evaluation feedback. This achieves deep coupling between simulation and optimization.

[0195] Multi-objective optimization engine: Employs population evolution-based heuristic algorithms, such as NSGA-II.

[0196] Initialization: An initial population containing multiple candidate river management schemes is generated randomly or empirically. Each scheme consists of a defined vector of decision variables. This indicates that a plan might include the boundary coordinates of the dredging area, the target dredging depth, the selected equipment model, the start date, and the dosage of degradation agent to be applied.

[0197] Fitness assessment: The plan The decision variable information (dredging parameters, degradation agent parameters, etc.) contained therein is input into the constructed watershed process simulation system as the "engineering intervention" input of the model.

[0198] Launch the simulation system and run a simulation of the river's environmental evolution after the implementation of the plan. The simulation needs to consider the hydrodynamic changes, sediment transport and redistribution, pollutant migration and transformation, riverbed erosion and deposition changes, and ecological responses under the intervention of the plan. The simulation timescale should be long enough to reflect the short-term effects of the plan (such as disturbance during dredging and pollution release) and long-term effects (such as sedimentation recovery rate, final pollutant concentration, and ecological restoration status).

[0199] Extract the information needed to calculate the multi-objective function from the output of the simulation system, such as the actual dredging volume in different areas, pollutant concentration at specific time points and locations, water depth, suspended sediment concentration, biodiversity index, etc.

[0200] Based on the extracted information, calculate the solution. In multiple objective function values .

[0201] Meanwhile, inspection plan Does the simulation satisfy all defined constraints? If the simulation results show that any constraint is violated (e.g., the predicted suspended sediment concentration at the downstream section exceeds the standard), the scheme is deemed infeasible.

[0202] Selection: A multi-objective selection operation is performed based on the multiple objective function values ​​for each scheme. NSGA-II uses non-dominated ranking to categorize schemes into different ranks, with higher ranks indicating better performance in the multi-objective space (less dominated by other schemes). Within the same rank, crowding distance is used to measure the sparsity of individuals, prioritizing individuals with higher crowding to maintain population diversity. The selection process retains superior schemes for the next generation.

[0203] Crossover (recombination) and mutation: Crossover and mutation operations are performed on the selected parent schemes to generate new offspring schemes. These operations are performed on the decision variable vector. This is done at the level of [the relevant department / entity]. For example, the boundary information of the dredging area of ​​two schemes can be cross-combined, or the dosage parameter of the degradation agent in a certain scheme can be randomly varied.

[0204] Termination Condition: Repeat the iterative processes of fitness evaluation, selection, crossover, and mutation until a preset termination condition is met. Output the non-dominated solution set from the external archive as the final set of optimized solutions (Pareto front). These solutions represent the optimal solutions that balance different objectives.

[0205] Dynamically adjusting the optimization engine's search strategy based on simulation feedback: Based on the results of simulation evaluation, dynamically adjust how it generates and selects new candidate solutions. The specific mechanism includes at least the following:

[0206] Adaptive Operator Probability Adjustment: During the optimization process, the algorithm records and analyzes the "success rate" or "contribution" of different crossover and mutation operators in generating excellent solutions (e.g., solutions that can enter higher non-dominated layers or fill sparse regions of the Pareto front). If, in recent iterations, a particular crossover operation (e.g., swapping the operation sequence of two dredging zones) or mutation operation (e.g., slightly adjusting the degradation agent dosage) more frequently produces better solutions, then in subsequent iterations, the algorithm dynamically increases the probability of using that operator or operators with similar parameter settings. Conversely, the probability of using poorly performing operators decreases. This allows the algorithm to automatically learn which parameter changes and combinations are more likely to produce improved solutions.

[0207] Search bias adjustment based on solution density and distribution: During the optimization process, the algorithm analyzes in real time the distribution of non-dominated solutions (points on the Pareto front) in the current external archive or population within the target space (a multidimensional space composed of defined objective functions). If solutions are very dense in certain regions, it indicates that the algorithm has fully explored those regions, and the probability of finding better solutions nearby is relatively low. In this case, when selecting parents or generating offspring, the algorithm reduces its bias towards searching these dense regions, instead encouraging the exploration of sparsely populated regions in the target space. This can be achieved by adjusting selection pressure, using niche techniques, or crowding distance. This enhances the algorithm's search capability across the entire Pareto front, resulting in a more evenly distributed and diverse set of optimization solutions.

[0208] Co-evolution with Surrogate Models: Since the computation of watershed process simulation systems is usually very time-consuming, directly feeding all candidate solutions into the simulation would greatly reduce optimization efficiency. To accelerate the optimization process, surrogate models can be constructed to approximate the evaluation function of the simulation system.

[0209] S6. Implementation, Dynamic Monitoring and Adaptive Adjustment: Compare monitoring data with model predictions. If significant deviations or new situations are found, the model will be updated and the plan will be re-optimized.

[0210] Organize the construction project according to the specific details of the generated preferred scheme (dredging area, depth, equipment, timing, degradation agent scheme, etc.).

[0211] Dynamic monitoring: Continuous monitoring of the river environment should be conducted before, during, and after dredging operations, as well as during the recovery period. Monitoring indicators should cover the key indicators collected in S1, and additional specific monitoring should be added for disturbances during construction and the effects of degradation agents. For example, turbidity sensors, DO sensors, suspended sediment sampling points, and pollutant concentration sampling points should be deployed in the dredging area and downstream; changes in sediment deposition should be monitored; the recovery of aquatic organisms should be assessed; and the concentrations of degradation agent parent material and metabolites should be monitored. The frequency and spatial density of monitoring data should be determined based on the operational stage and sensitivity.

[0212] Adaptive adjustment:

[0213] Comparative analysis of monitoring results and model predictions: The acquired real-time and phased monitoring data are compared with the prediction results of the watershed process simulation system. For example, the measured suspended sediment concentration curves at specific cross sections are compared with the simulated prediction curves, and the dissolved oxygen (DO) recovery rate in dredged areas is compared with the simulated predicted rate.

[0214] Trigger Model Update: If the monitoring results deviate significantly from the model predictions (exceeding the preset threshold), or if new changes occur in the river system that cannot be explained by the current structure or parameters of the model (such as sudden pollution events or extreme hydrological conditions), then the parameters or structure of the watershed process simulation system will be updated.

[0215] Online correction for specific physical parameters: For deviations that can be improved in simulation accuracy through parameter adjustments, data assimilation techniques (such as Kalman filtering, ensemble Kalman filtering, particle filtering, and variational assimilation) are used to integrate real-time monitoring data into the simulation system, enabling online inversion and updating of key physical parameter values. For example, if the simulation underestimates the diffusion rate of suspended sediment, the sediment dispersion coefficient may need to be adjusted; if dissolved oxygen (DO) recovery is slower than predicted, the reaeration coefficient may need to be reduced or the sediment oxygen consumption rate increased.

[0216] Model Structure Diagnosis and Adjustment Based on Error Propagation Analysis: If the phenomena shown by monitoring data differ fundamentally from model predictions (e.g., prediction of sedimentation in a region, but actual erosion), or if the errors exhibit a systematic and large-scale distribution, it may indicate a defect in the model structure. Error propagation analysis, combined with parameter sensitivity analysis, is conducted to diagnose which model assumptions (e.g., two-dimensional assumptions are not applicable to strong vertical flows), simplified governing equations (e.g., failure to consider sediment consolidation effects), or missing subprocesses (e.g., failure to include specific reaction pathways for a particular pollutant) are the main causes of the errors. Based on the diagnostic results, the model structure is adjusted, for example, by upgrading the two-dimensional hydrodynamic model to a three-dimensional model, or by adding new reaction kinetic equations to the pollutant module.

[0217] Re-running the optimization process: After model updates, if it is determined that the current implementation plan may no longer be optimal under new environmental conditions, the multi-objective optimization process in step S5 is selectively re-run. The updated watershed process simulation system is used as the evaluation module, and the latest environmental conditions may be incorporated as the starting point for optimization. Re-optimization will generate adjustment plans adapted to the new environmental conditions. For example, if monitoring reveals unexpected severe siltation downstream of the dredging area, the adjustment plan may include increasing local dredging downstream or changing the dredging sequence.

[0218] Plan Adjustment and Implementation: Based on the adjusted plan obtained through re-optimization, corresponding modifications and adjustments will be made to the ongoing dredging operations and the application of degradation agents.

[0219] To fully verify the effectiveness of the method of this invention, we will conduct the following experimental design. By comparing the performance of different schemes in actual river environments, we will demonstrate that the scheme based on dynamic simulation and optimization decision-making has better overall benefits and environmental adaptability than traditional methods and methods based only on preliminary scheme evaluation.

[0220] Experimental Site: A section of a river with typical siltation and a certain degree of water pollution was selected as the experimental area. This area should have relatively complete historical environmental monitoring data to facilitate model calibration and validation.

[0221] Control group setup:

[0222] Control Group 1 (Traditional Scheme): Based on engineering experience and simplified calculations, this scheme considers only local indicators such as the siltation thickness of the river section to be dredged, and formulates a traditional dredging plan. Parameters such as dredging equipment, scope, depth, and timing are determined based on experience. No watershed-scale simulation or multi-objective optimization is performed.

[0223] Control Group 2 (based on preliminary scheme evaluation): This approach uses existing technology, where engineers first develop a preliminary dredging plan, which is then evaluated using a watershed-scale model. Based on the evaluation results, one or a limited number of scheme revisions may be made, but no systematic optimization search of a large number of potential schemes is conducted, nor is adaptive adjustment based on dynamic monitoring performed.

[0224] Experimental Group (Invention Scheme): Following the method and flow of this invention, a watershed process simulation system is constructed and calibrated through multi-source data acquisition and feature analysis. Scheme elements are defined parametrically, a knowledge base is built, and a multi-objective optimization problem is formally defined. A set of Pareto optimal schemes is generated using a simulation feedback-based multi-objective optimization engine. One or more representative schemes are selected from this set of optimal schemes for implementation. Continuous monitoring is conducted during and after implementation, and the model is adaptively updated and the scheme is dynamically adjusted based on the monitoring results.

[0225] Performance testing methods:

[0226] Dredging effect: After the dredging operation was completed, the actual riverbed elevation of the dredged area was obtained through high-precision underwater topographic surveying, and the actual dredged volume was calculated. The dredging completion rate (the ratio of actual dredged volume to target dredged volume) of the experimental and control groups was compared. Simultaneously, the siltation rate in the dredged area and a specific downstream section was monitored to assess the sustainability of the dredging effect.

[0227] Pollutant Reduction Effect: Target pollutants were sampled and analyzed from the sediment and water in the dredged area at different time points before and after dredging. The total pollutant reduction rate in the dredged area and downstream waters was calculated after implementing different schemes. For schemes using degradation agents, changes in pollutant concentrations and degradation products in the agent's effective area were monitored.

[0228] Economic Costs: Record and statistically analyze the total actual investment required for implementing different schemes, including initial investment, dredging operation costs, transportation and disposal costs, degradation agent costs, monitoring costs, etc. Compare the actual total costs of the experimental group and the control group.

[0229] Environmental impact:

[0230] Water quality: Monitor changes in key water quality parameters such as suspended solids, dissolved oxygen, and turbidity at sensitive sections downstream of the dredging area during and after the dredging operation to assess the degree of short-term water quality disturbance.

[0231] Ecology: Before and during the recovery period after dredging, benthic organisms and aquatic vegetation in the experimental and control areas were investigated to assess the ecosystem's recovery status, such as changes in biodiversity indices. The potential impact of the degradation agent on non-target organisms was monitored.

[0232] Flood discharge capacity and navigation guarantee: After dredging is completed, the changes in cross-sectional area and roughness of the dredged area will be assessed through hydraulic simulation or field measurements to calculate the improvement in flood discharge capacity. For waterway areas, the actual water depth after dredging will be monitored to ensure that navigation requirements are met.

[0233] Adaptability of the plan: If unexpected environmental changes (such as major floods or sudden pollution) occur during or after the experiment, record the response capabilities and adjustment processes of different plans. Evaluate the effectiveness of the experimental group in handling emergencies based on dynamic monitoring and adaptive adjustment mechanisms.

[0234] Data Analysis and Comparison: Statistical analysis and visualization of the acquired performance test data were performed. Multi-objective evaluation methods (such as TOPSIS, VIKOR, etc.) were used to quantitatively compare the overall performance of the experimental and control groups, analyzing the advantages of the proposed solution in multi-objective optimization. Simultaneously, the adjustment efficiency and effectiveness of the experimental group's solution in responding to environmental changes were compared and analyzed.

[0235] Dredging Scope and Depth: The dredging scope is typically limited to severely silted sections of the river, while the depth can be set with a reasonable upper limit, such as based on the historical deepest riverbed elevation or the design depth of the navigation channel. For example, areas with siltation thickness greater than 1 meter are considered potential dredging areas, with a maximum dredging depth of 3 meters.

[0236] Dredging Equipment Types and Combinations: The performance parameters of the equipment in the knowledge base determine its applicable scope. For example, cutter suction dredgers are suitable for viscous silt and have limitations on maximum operating depth. The optimization algorithm will select and combine appropriate equipment types.

[0237] Dredged material treatment methods: The available disposal sites and resource recovery processes are defined in the knowledge base, and optimization will select the treatment combination with the lowest cost, sufficient capacity and minimal environmental impact.

[0238] Application of pollutant degradation agents:

[0239] Type and Ratio: Based on the pollutant type determined in step S1, select the target degrading agent from the knowledge base. If a compound is required, the proportions of each component can vary within a certain range, and the optimization algorithm will explore the optimal ratio. For example, for a certain organic matter degrading microbial agent, the ratio of microbial strain A to microbial strain B can be optimized between 3:1 and 1:3.

[0240] Application location and dosage: Location is typically limited to areas of contaminated sediment or contaminated water. Dosage can be adjusted based on the recommended dosage, for example, by searching for an optimal value between 0.5 and 2 times the recommended dosage. Application frequency can also be selected from several common patterns or the interval between applications can be optimized.

[0241] Chemical-related technologies: Examples of pollutant degradation agent application schemes.

[0242] Pollutant degradation agent application plan (as part of the river siltation treatment plan)

[0243] Target pollutants: Assuming that the S1 data analysis determines that the sediment in the riverbed to be treated contains high concentrations of organic pollutants and heavy metal Cd.

[0244] Degradable Agent Knowledge Base: The knowledge base includes:

[0245] Organic matter degradation compound microbial agent A: mainly targets organic matter, suitable for pH 6-8, temperature 15-35℃, recommended dosage 1kg / 100m² bottom sediment surface area, cost 50 yuan / kg.

[0246] Cd passivating agent B: The main components are phosphates or lime-like substances. It fixes Cd through precipitation or adsorption. It is suitable for pH 7-9. The recommended dosage is 0.5-1 kg / m² of sediment volume. The cost is 30 yuan / kg.

[0247] Compound synergist C: can improve the activity of bacterial agent A, with a recommended ratio of 10-20% of bacterial agent A, costing 80 yuan / kg.

[0248] Parameterized definition:

[0249] Whether to apply bacterial agent A, passivating agent B, or synergist C (Boolean variable).

[0250] If applied, the mass ratio of microbial agent A to synergist C.

[0251] Spatial extent of the application area (potential application areas are determined based on pollution concentration distribution maps).

[0252] The dosage of fungicide A per unit area (kg / m²) for each application area.

[0253] The application dosage of passivating agent B per unit volume in each application area (kg / m³ bottom mud).

[0254] Application time (before dredging, during dredging, after dredging).

[0255] Application frequency (single application, once a week, once every two weeks).

[0256] Watershed process simulation system: The pollutant migration and transformation module needs to include the biodegradation kinetics equations of organic matter, the adsorption-desorption and precipitation / dissolution equilibrium equations of Cd in the water-sludge system, and the simulation of the migration and diffusion of the degrading agent in the environment and possible side reactions. The degradation rate and passivation effect are related to the dosage of the degrading agent, environmental conditions (pH, temperature), and sediment properties.

[0257] Multi-objective optimization: The objective function includes maximizing the benefits of pollutant reduction (organic matter and Cd) and minimizing the total economic cost. Constraints include that the concentration of Cd in water bodies is below environmental standards and that the toxicity of the degradation agent to non-target organisms after application is below the safety threshold.

[0258] Example 2: Degrading agent application scheme optimized based on the present invention

[0259] Through the optimization process in steps S1-S5, the optimization engine, combined with the simulation system evaluation, generated a set of Pareto optimal solutions. One of the preferred solutions is:

[0260] Dredging scope and depth: Specific siltation area, depth 2m.

[0261] Dredging schedule: to be carried out during the dry season.

[0262] Dredging equipment: Low-disturbance cutter suction dredger.

[0263] Dredged material treatment: transported to a designated disposal site and solidified.

[0264] Degrader application procedure:

[0265] Apply microbial agent A, passivating agent B, and synergist C.

[0266] The ratio of bacterial agent A to synergist C is 5:1.

[0267] Application area: Substrate areas with high organic matter and high Cd concentration.

[0268] Dosage of microbial agent A: 1.2 kg / 100 m² for highly polluted areas and 0.8 kg / 100 m² for moderately polluted areas.

[0269] Passivating agent B application dosage: 0.7 kg / m³ of bottom sediment in areas with high Cd concentration.

[0270] Application time: Begin two weeks after dredging is completed.

[0271] Application frequency: Apply bacterial agent A and synergist C once a week for a total of 4 times; apply passivator B once.

[0272] Example 3: Endpoint values ​​and range of values ​​in the optimization scheme

[0273] To further demonstrate the search capabilities of the optimization algorithm, we selected two other solutions generated during the S5 optimization process. These solutions may be located at different positions on the Pareto front, representing trade-offs between different objectives.

[0274] Example 3a (Focusing on pollutant reduction, high cost):

[0275] Application scheme for degradation agents: The dosage of microbial agent A is uniformly 1.5 kg / 100 m², and the dosage of passivating agent B is uniformly 1 kg / m³ of sediment. The ratio of microbial agent A to synergist C is 4:1. Application frequency: Microbial agent A and synergist C are applied twice a week for a total of 6 times; passivating agent B is applied once. The application area is larger, covering moderately polluted areas.

[0276] Example 3b (focusing on cost control, with slightly inferior results):

[0277] Application scheme for degradation agents: Apply only bacterial agent A and passivating agent B, without applying synergist C. The application dosage of bacterial agent A is uniformly 0.7 kg / 100 m², and the application dosage of passivating agent B is uniformly 0.5 kg / m³ of sediment. Application frequency: Bacterial agent A is applied once a week for a total of 3 times; passivating agent B is applied once. Application area is limited to the most contaminated area.

[0278] Scale settings:

[0279] To compare the advantages of the optimized solution of the present invention, the following comparative examples are provided:

[0280] Comparative Example 1 (Traditional dredging, no degradation agent): Only dredging was carried out, and the dredging range and depth were determined based on traditional methods. No pollutant degradation agent was applied.

[0281] Comparative Example 2 (Dredging + Empirical Degrading Agent Application): The same dredging process as Comparative Example 1 was performed. The degrading agent application scheme was based on expert experience and the general dosage and frequency recommended by the degrading agent manufacturer. For example, the dosage of microbial agent A was uniformly 1 kg / 100 m², and the dosage of passivating agent B was uniformly 0.8 kg / m³ of bottom sediment, applied in a single application. Differences in different contaminated areas and optimal timing were not considered.

[0282] Comparative experiments and results analysis:

[0283] In the selected experimental river section, the schemes of Example 2, Example 3a, Example 3b, Comparative Example 1, and Comparative Example 2 were implemented simultaneously. After dredging and at different time points (e.g., 1 month, 3 months, and 6 months) after the application of the degradation agent, the sediment and water pollutants (organic matter, Cd) were sampled and analyzed, and the total cost was recorded.

[0284] Comparison table example:

[0285] Solution Name Dredging completion rate (%) Organic matter reduction rate (%) Cd reduction rate (%) Total economic cost (ten thousand yuan) Downstream water quality disturbance (short-term) Ecological restoration potential Remark Comparative Example 1 (Traditional Dredging) 85 20 15 400 middle lower Only physical removal of some polluted sediment Comparative Example 2 (Dredging + Experienced Application) 85 35 30 450 medium middle The effectiveness of degradation agents is limited by non-optimal solutions. Example 3b (Cost Control) 88 40 38 480 medium medium Lower cost, but limited improvement in effectiveness. Example 3a (Effect Priority) 92 65 60 650 lower higher The effect is significant, but the cost is high. Example 2 (Optimization of the Invention) 90 55 52 520 lower higher Achieving a better balance between cost and effectiveness with minimal disturbance

[0286] Analysis: Comparative Example 1 only dredged the sediment, which had limited effect on dissolved pollutants and residual sediment pollution.

[0287] Comparative Example 2 added a degradation agent, which improved the effect, but since it was an empirical approach, the potential of the degradation agent was not fully utilized.

[0288] Examples 3b and 3a illustrate the solutions generated by the optimization algorithm under different objective preferences. Example 3a, in pursuit of better results, incurs significantly higher costs than Example 2. Example 3b controls costs, but its performance is not as good as Example 2.

[0289] Example 2, as an optimized scheme generated by the method of the present invention, typically achieves a better overall balance in terms of dredging completion, pollutant reduction rate, economic cost, environmental impact (lower predicted disturbance), and ecological restoration potential compared to the comparative examples and those focusing on extremes. This verifies that the method of the present invention, through multi-objective optimization and simulation evaluation, can generate river silt treatment schemes with higher overall benefits.

[0290] Further monitoring and adaptive adjustments will be made based on actual feedback to optimize the solution and ensure its continued effectiveness.

[0291] The above is a detailed extension and description of the embodiments of the method of the present invention. The above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Ordinary variations and substitutions made by those skilled in the art within the scope of the technical solutions of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for river siltation control based on dynamic simulation and optimization decision-making, characterized in that, Includes the following steps: S1. Multi-source environmental data acquisition and feature analysis: Acquire and process historical and real-time multi-dimensional environmental data of the river to be treated and its affected areas. The data should include at least topography, hydrology and hydrodynamics, sediment characteristics, water quality and pollutant distribution, and ecosystem status data. The data is fused and processed to extract key spatiotemporal features that characterize river siltation, water pollution, and ecological health. S2. Construction and Validation of Watershed Process Simulation System: Construct a watershed-scale coupled watershed process simulation system capable of simulating river hydrodynamics, sediment transport, pollutant migration and transformation, and ecological response processes; The key features extracted in step S1 and historical monitoring data are used to calibrate and verify the parameters of the watershed process simulation system to ensure its accuracy in simulating environmental processes. S3. Parameterization and knowledge base support of river management scheme elements: Parameterization defines the decision variables that constitute the river management scheme. The decision variables include at least the dredging scope, dredging depth, dredging sequence, dredging equipment type and combination, dredged material treatment method, and pollutant degradation agent type, formula ratio, application location and dosage. Construct a knowledge base that includes the relevant performance, cost, environmental impact, and applicable conditions of the above elements; S4. Formal Definition of Multi-Objective Optimization Problem: Define multiple objective functions to evaluate the comprehensive benefits of river management schemes. The objective functions should at least cover dredging efficiency, pollutant reduction rate, total project cost, environmental impact minimization, and ecological restoration effect. Define the technical, economic, environmental, and regulatory constraints that must be met during the operation. S5. Multi-objective optimization engine driven by simulation feedback and scheme generation: A multi-objective optimization engine is adopted, and the watershed process simulation system is embedded into the optimization iteration loop as its core evaluation module; The multi-objective optimization engine defines decision variables and multi-objective optimization problems. During its search process, it calls the watershed process simulation system in real time to simulate the environmental state evolution and expected results of multi-objective evaluation indicators after the implementation of different candidate governance schemes. Based on the prediction feedback of the simulation system, it dynamically adjusts its search strategy to generate a set of river governance schemes that meet the constraints and make multiple objective functions optimal. S6. Implementation, Dynamic Monitoring and Adaptive Adjustment: Implement the scheme generated in step S5, and continuously monitor changes in environmental conditions during and after implementation; compare and analyze the monitoring results with the prediction results of the watershed process simulation system; if significant deviations or new environmental disturbances occur, trigger the updating of the parameters or structure of the watershed process simulation system, and selectively rerun the optimization process of step S5 to generate an adjustment scheme adapted to the new environmental conditions.

2. The river siltation treatment method based on dynamic simulation and optimization decision-making according to claim 1, characterized in that, The watershed process simulation system integrates at least the following sub-models: Hydrodynamics Module: Based on the principles of fluid mass conservation and momentum conservation, this module uses two-dimensional or three-dimensional numerical methods to solve the governing equations to simulate the hydrodynamic processes of water level, velocity field, and flow distribution in rivers and related water areas driven by water flow, tides, and wind fields under the conditions of river topographic boundaries and riverbed roughness. Sediment transport module: This module is bidirectionally coupled with the hydrodynamic module. Based on the flow conditions and physical properties of sediment, including particle size distribution, density, and settling velocity, it simulates the convection, diffusion, and settling processes of suspended sediment and the initiation and transport processes of bedload sediment, and considers the composition of riverbed sediment and the influence of critical shear stress. Riverbed Evolution Module: This module is based on the principle of sediment continuity and dynamically updates the riverbed topography and elevation based on the amount of sediment erosion and deposition per unit time calculated by the sediment transport module.

3. The river siltation treatment method based on dynamic simulation and optimization decision-making according to claim 2, characterized in that, In the sediment transport module, the calculation of the erosion rate and deposition rate of riverbed sediment is further limited to: The erosion rate is calculated based on the comparison between the actual shear stress of the water flow acting on the riverbed and the critical erosion shear stress of the sediment. When the actual shear stress is greater than the critical erosion shear stress, sediment erosion occurs. The erosion rate is related to the difference or ratio between the two and the properties of the sediment. The sedimentation rate is calculated based on the sedimentation characteristics of the sediment and the sediment concentration in the near-bed flow. It can also take into account the comparison between the actual shear stress of the flow and the critical sedimentation shear stress of the sediment. When the actual shear stress is less than the critical sedimentation shear stress, it is conducive to sediment deposition.

4. The river siltation treatment method based on dynamic simulation and optimization decision-making according to claim 1, characterized in that, The watershed process simulation system further integrates a pollutant migration and transformation module, which is coupled with the hydrodynamics module and the sediment transport module to simulate the following processes of one or more target pollutants in the study water area: The distribution of pollutants between the dissolved phase and the particulate phase, including adsorption and desorption processes; Convective transport of dissolved and adsorbed pollutants with water flow and diffuse transport caused by concentration gradients; Adsorbed pollutants are deposited on the riverbed as suspended sediment settles, and pollutants already deposited on the riverbed are re-entered into the water as bottom sediment is resuspended. The biochemical degradation or transformation process of pollutants themselves, the rate of which is affected by the nature of the pollutants, environmental conditions and microbial population factors.

5. The river siltation treatment method based on dynamic simulation and optimization decision-making according to claim 1, characterized in that, The parameterized definition of the elements of the river management plan includes: Spatial parameters for dredging: These include a sequence of geographic coordinate points or a set of grid cells that define the boundaries of the dredging area, as well as the target post-dredging riverbed elevation or planned dredging thickness set within each area or cell; Dredging time parameters: including the planned start date of each dredging area, the expected number of days or hours of continuous operation, and the logical relationship between the operation sequence of different areas; Dredging process and equipment parameters: Model index of dredging vessels selected from the knowledge base, required quantity and key operational parameters; Dredged material disposal and resource utilization parameters: selected dredged material transfer or final disposal site number, transportation distance, acceptable dredged material characteristics, and selected resource utilization process type and its corresponding processing capacity and cost coefficient; Pollutant degradation agent application parameters: Degradation agent selection and compatibility: Identifiers for biological or chemical agents targeting the pollutant; if it is a compound, it includes the precise mass or volume percentage of each component. Spatial application strategy: The application area of ​​the degradative agent is defined in the form of a two-dimensional grid, and the spatial distribution function of the planned application dose or total application amount of each grid cell; Application time strategy: the start time, duration and frequency of application of the degradation agent. The application frequency includes single shock dose, multiple pulse dose or continuous low dose.

6. The river siltation treatment method based on dynamic simulation and optimization decision-making according to claim 1, wherein the objective function of the multi-objective optimization problem has a specific mathematical expression or its constituent terms including at least: Maximize the completion rate of dredging work: , in It is a plan The predicted actual dredging volume for the i-th dredging area. It is the target dredging volume. This represents the total number of dredged areas. These are weighting coefficients; Maximizing the benefits of pollutant reduction: , in It is the initial total amount of the j-th target pollutant. It is a plan The final total amount of the pollutant predicted after implementation, It is the number of target pollutant types. These are the weighting or benefit conversion factors for the reduction of different pollutants; Minimize total economic cost: The costs are fixed costs, dredging operation costs, transportation costs, disposal costs, degradation agent costs, and monitoring costs, minus the potential revenue from the resource utilization of dredged materials. Minimize environmental risks: , in It is a plan The predicted k-th negative environmental impact indicator includes the peak concentration of suspended solids at the downstream section, the short-term water quality deterioration index caused by dredging disturbance, and the risk of exceeding the standard concentration of degradation agent residue or by-products. It refers to the number of environmental impact indicators. It is weight; Maximizing the restoration of ecological functions: , in It is a comprehensive ecological evaluation function whose inputs are multiple predicted ecological indicators, such as improved water transparency, increased dissolved oxygen levels, potential for recovery of benthic biodiversity index, and increased coverage of indicator aquatic plants.

7. The river siltation treatment method based on dynamic simulation and optimization decision-making according to claim 1, characterized in that, The constraints of the multi-objective optimization problem, in their specific mathematical expression or logical judgment, include at least the following: Budget ceiling: Total economic cost ; Overall project timeline constraints: The operation time and equipment dispatch time in each area were taken into account. Channel function assurance: For any point (x, y) within the channel area, the water depth after dredging ; Environmental quality standard compliance: For all pre-set environmental control points m, the predicted concentrations of all key water quality parameters q during the operation and recovery periods. ; Ecological red line area protection: areas where dredging or degradation agents are applied. and ecological protection red line areas There is an intersection Then it requires Or the intensity of operations / disturbance level within this intersection area. ; Dredging disposal site capacity: Scheme Total volume of dredged material produced ; Degrader safety threshold: The predicted concentration of the parent degrader or its known major degradation intermediates at any point in the environment. ; Riverbank or levee stability: The minimum safety factor for adjacent riverbanks or levees after dredging. .

8. The river siltation treatment method based on dynamic simulation and optimization decision-making according to claim 1, characterized in that, The multi-objective optimization engine is based on a population evolution heuristic algorithm. This algorithm maintains a population of candidate river management schemes and iteratively improves the quality of the schemes in the population by repeatedly applying selection, crossover, and mutation evolution operators. Specifically, the implementation is as follows: Initialization: An initial population containing multiple different river management schemes is generated randomly or according to empirical rules. Each scheme is defined by a vector of decision variables. express; Fitness assessment: For each individual in the population, a watershed process simulation system is invoked to predict the environmental evolution process after the implementation of the scheme, and multiple objective function values ​​are calculated based on this prediction. ; Selection: Based on the individual's objective function value, non-dominated sorting and crowding calculation are used to select superior individuals from the current population to enter the next generation or to participate in reproduction as parents. Crossover: Select parent pairs of individuals and their decision variable vectors. The crossover operation involves exchanging or combining some or all of the elements to generate new offspring individuals. The design of the crossover operation should take into account the characteristics of the decision variables. Variation: The vector of decision variables for offspring individuals Small random perturbations or changes are made to certain elements in order to maintain population diversity and explore new regions of the solution space; Elite strategy: Store the non-dominated solutions generated in each generation in an external archive and ensure that these solutions can be passed on to the next generation to prevent the loss of excellent solutions; Termination conditions: When the preset maximum number of iterations, computation time limit, and the quality of solutions in the population no longer improve, the termination condition is reached.

9. The river siltation treatment method based on dynamic simulation and optimization decision-making according to claim 1, characterized in that, The search strategy for dynamically adjusting a multi-objective optimization engine based on simulated feedback includes at least the following: Adaptive operator probability adjustment: Records the frequency or utility of the algorithm in the past few generations in successfully producing better solutions through crossover and mutation operators. If a certain operator or its parameter settings have performed better recently, the probability of its selection is dynamically increased or its parameter range is adjusted; conversely, it is decreased. Search bias adjustment based on solution density and distribution: Real-time analysis of the distribution density of current Pareto front solutions in the target space. If solutions are too dense in certain regions, the priority of exploring these regions is reduced when selecting a parent or generating a new individual; co-evolution with the surrogate model.

10. A method for river siltation control based on dynamic simulation and optimization decision-making according to claim 1, characterized in that, The updating of the parameters or structure of the watershed process simulation system includes: For online correction of physical parameters: Kalman filtering is used to compare the acquired real-time monitoring data, including water level, flow velocity, measured suspended sediment concentration, and key pollutant concentration at the cross-section, with the corresponding outputs of the simulation system. By minimizing the deviation between the two, key physical parameters in the simulation system are retrieved and updated online, such as the riverbed roughness coefficient and the critical initiation shear stress of sediment. Comprehensive degradation rate coefficient of pollutants or adsorption partition coefficient ; Model structure diagnosis and adjustment based on error propagation analysis: Analyze the temporal and spatial distribution characteristics of the prediction error of the simulation system, and combine with sensitivity analysis to identify which model assumptions, simplified control equations, or missing subprocesses may be the main sources of error; If the diagnosis indicates that the existing model structure cannot adequately describe the observed phenomena, the model structure is adjusted, including upgrading from a two-dimensional model to a three-dimensional model, or introducing new reaction kinetic equations or coupling new ecological modules into the pollutant migration and transformation module.

Images