A pollution collaborative control optimization method and system based on a watershed system

Abstract

The application discloses a watershed system-based pollution collaborative control optimization method and system, relates to the field of watershed environmental management and pollution control, and comprises the following steps: setting a pollutant concentration target of a watershed outlet, and collecting hydrological and water quality data. By analyzing the section relationship of the trunk stream and the branch stream, a contribution coefficient method or a model prediction method is selected according to the linear degree, and a quantitative response relationship between the upstream pollution load and the outlet water quality is established. A multi-objective algorithm is used to obtain an optimal allocation scheme set of the pollution load by taking the fairness of burden reduction and the total quantity efficiency as targets. Based on the scheme characteristics of different water conditions, and the social economy and engineering feasibility, a control scheme and a concentration control target are selected, which give consideration to the fairness and the efficiency. The application realizes precise total nitrogen emission reduction and sea inlet concentration control at the watershed scale, adapts to complex water conditions of large watersheds, solves the problem that the existing control target backstepping method is not applicable to large watersheds, and improves the systematicness and scientificity of total nitrogen pollution treatment of large watersheds.

Description

Technical Field

[0001] This application relates to the field of watershed environmental management and pollution control, and in particular to a pollution collaborative control optimization method and system based on watershed systems. Background Technology

[0002] Rivers are crucial links connecting land and sea, and the input of land-based pollutants is a major cause of eutrophication in nearshore waters. Total nitrogen (TN) pollution, in particular, involves complex migration and transformation processes, including the inflow of different tributaries within the watershed, degradation along the river, and retention. Traditional watershed management often adopts a "localized treatment" model, setting standards for individual river sections, lacking a systematic consideration from "mountain top to sea." The key to controlling TN concentrations entering the sea is how to use downstream pollution concentration control targets as a guide to deduce upstream control targets. Currently, there are two main methods for deduce control targets: one is to use a one-dimensional water quality diffusion equation as the core, combined with water quality mixing models or water environmental capacity models, to deduce inflow concentrations or discharge volumes; the other is to couple multiple water quantity and quality models, such as QUAL2K and SWAT (Soil and Water Assessment Tool), to find scenarios that allow downstream sections to meet standards through repeated simulations. Both methods are suitable for small rivers and watersheds with relatively stable hydrology and water quality, but not for large watersheds with complex hydrological conditions. Furthermore, controlling the total nitrogen concentration entering the sea from a large watershed requires simultaneous regulation at multiple cross-sections. Coordinating the relationships between these cross-sections and finding the optimal control scheme is also a key consideration. Currently, multi-objective optimization algorithms are widely used in environmental fields such as water quality monitoring and water resource allocation. Through multi-objective optimization of reservoirs or water resources, water quality optimization can be achieved. While water quality is often incorporated into multi-objective systems as a single objective, spatial optimization of water quality at multiple cross-sections is rarely considered.

[0003] Currently, given the complex hydrological conditions in large river basins, the inapplicability of existing control target back-calculation methods, and the urgent need for multi-section coordinated regulation but lack of consideration for multi-section water quality spatial optimization, existing river basin water quality management methods suffer from the following main problems: First, they lack a systematic consideration of the river basin, often focusing only on achieving standards at a single section while neglecting the linkage between upstream and outlet targets, resulting in upstream reduction measures failing to effectively respond to the environmental capacity requirements of estuaries and seas; second, there is a disconnect between model simulation and optimization decision-making, as mechanistic models are computationally time-consuming and difficult to directly apply to multi-objective optimization processes requiring thousands of iterations; while purely data-driven models lack physical mechanism explanations and are susceptible to data gaps; third, allocation schemes are simplistic: when formulating reduction tasks, it is difficult to find a balance between "maximizing environmental benefits (efficiency)" and "regional development rights (equity)," lacking scientific quantitative tools for weighing these factors. Therefore, there is an urgent need for a coordinated control method that can couple hydrological process simulation, rapid response models, and multi-objective optimization to achieve precise emission reduction at the river basin scale. Summary of the Invention

[0004] The purpose of this application is to provide a pollution synergistic control optimization method and system based on watershed systems, which can achieve precise reduction of total nitrogen emissions and control of sea discharge concentration at the watershed scale, adapt to the complex hydrological conditions of large watersheds, solve the problem that existing control target back-inference methods are not applicable to large watersheds, and improve the systematicness and scientific nature of total nitrogen pollution control in large watersheds.

[0005] To achieve the above objectives, this application provides the following solution: Firstly, this application provides a pollution collaborative control optimization method based on a watershed system. The method includes: setting a target concentration of the pollutant to be controlled at the watershed outlet within a study area; the study area being located within a target watershed system; collecting hydrological and water quality monitoring data of the study area and supplementing default values ​​by constructing and calibrating a SWAT hydrological and water quality model; establishing the spatial response relationship of watershed water quality in the study area based on the target concentration, specifically including: determining the corresponding relationship between the main stream and tributary sections in the study area based on the target concentration and verifying the linear relationship between each section; determining whether the degree of linear relationship is greater than a relationship threshold; if so, executing the contribution coefficient method, calculating the load proportion coefficient between adjacent sections to infer the load contribution coefficient of each tributary to the watershed outlet; if not, executing the model prediction method, using a trained machine learning model to predict the upstream sections of each tributary and the main stream. The nonlinear mapping relationship between pollutant load and estuary section pollutant load yields the spatial response relationship of watershed water quality. The optimization objectives are to minimize the difference in the percentage reduction of total nitrogen load in each tributary and the total reduction in total nitrogen load in each tributary. Using the pollutant concentration or load control targets at the estuary section as constraints, a multi-objective optimization algorithm is constructed. Based on hydrological and water quality monitoring data and the spatial response relationship of watershed water quality, the algorithm iteratively optimizes the allowable discharge or reduction ratio of pollutants at each control section to obtain the Pareto optimal solution set. The Pareto front characteristics at different times within the Pareto optimal solution set are analyzed. Based on the socio-economic and engineering feasibility of the watershed, a control scheme for pollutants that balances fairness and environmental benefits / efficiency is selected from the Pareto optimal solution set, and the concentration control targets for pollutants at each section are determined.

[0006] In a second aspect, this application also provides a computer system, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the pollution collaborative control optimization method based on the watershed system described in the first aspect.

[0007] According to the specific embodiments provided in this application, the following technical effects are disclosed: This application establishes a target for pollutant concentration at the watershed outlet and integrates and supplements monitoring data using the SWAT hydrological model, providing a reliable foundation for analysis. Subsequently, an accurate and efficient spatial response relationship between upstream load and outlet water quality is established. Based on the degree of linearity between the main stream and tributary sections, either the contribution coefficient method or machine learning model prediction method is selected to accurately characterize the migration and transformation patterns of pollutants under complex hydrological conditions, overcoming the limitations of traditional methods in large watersheds. A multi-objective optimization algorithm is constructed, with the dual objectives of minimizing the difference in reduction percentage among tributaries and minimizing the total reduction, and with outlet compliance as a hard constraint. Iterative optimization using the algorithm generates a set of Pareto-optimal load allocation schemes that simultaneously consider environmental benefits and regional development interests. Finally, a comprehensive evaluation and selection process is conducted, combining the characteristics of schemes in different hydrological periods with socio-economic and engineering feasibility, to determine scientifically sound and implementable concentration control targets for each section. In summary, this application achieves precise reduction of total nitrogen emissions and control of its concentration in the sea at the watershed scale, adapts to the complex hydrological conditions of large watersheds, solves the problem that existing methods for back-calculating control targets are not applicable to large watersheds, and improves the systematicness and scientific nature of total nitrogen pollution control in large watersheds. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0009] Figure 1 This is a flowchart illustrating a pollution collaborative control optimization method based on a watershed system, provided in an embodiment of this application.

[0010] Figure 2 A schematic diagram of the middle and lower reaches of the Yellow River and its major tributaries provided for embodiments of this application.

[0011] Figure 3 This is a schematic diagram illustrating the calibration and verification results of the SWAT hydrological and water quality model provided in this application embodiment; wherein, Figure 3 In the diagram, 'a' represents the flow rate determination verification results at the Lijin section. Figure 3 In the diagram, b represents the verification results of the total nitrogen ratio determination at the Lijin section. Figure 3 In the figure, c is a schematic diagram of the flow rate determination verification results of the Fenhe River (Hejin) section; Figure 3 In the figure, d is a schematic diagram of the verification results of the total nitrogen rate determination of the Fenhe River (Hejin) section.

[0012] Figure 4 A schematic diagram of the Pareto optimal solution set provided in an embodiment of this application; wherein... Figure 4 In the diagram, 'a' represents the Pareto optimal solution set during the non-flood season. Figure 4In the diagram, b represents the Pareto optimal solution set during the flood season.

[0013] Figure 5 A schematic diagram comparing the percentage reduction of typical solutions during non-flood season and flood season provided in this application embodiment; wherein, Figure 5 In the diagram, 'a' represents the percentage reduction under a typical non-flood season plan. Figure 5 In the diagram, b represents the percentage reduction under a typical flood season plan.

[0014] Figure 6 This is a schematic diagram illustrating the total nitrogen concentration control target provided in an embodiment of this application; wherein, Figure 6 In the diagram, 'a' represents the total nitrogen concentration control target during the non-flood season. Figure 6 In the diagram, b represents the total nitrogen concentration control target during the flood season.

[0015] Figure 7 This is an internal structure diagram of a computer system provided in an embodiment of this application. Detailed Implementation

[0016] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0017] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0018] Example 1, as Figure 1 As shown in the figure, this embodiment provides a pollution collaborative control optimization method based on a watershed system, the method comprising: S1. Set the target concentration of the pollutant to be controlled at the outlet of the watershed within the study area; the study area is located within the target watershed system.

[0019] Furthermore, the scale of the target concentration includes at least one of the following: monthly scale, seasonal scale, and annual scale.

[0020] S2. Collect hydrological and water quality monitoring data of the study area and supplement the default values ​​by constructing and calibrating the SWAT hydrological and water quality model.

[0021] Further, step S2 specifically includes: determining the control sections of the main stream and major tributaries of the watershed in the study area, and collecting corresponding flow monitoring data and pollutant monitoring data. When there are areas or periods with insufficient monitoring data, the data collected are replaced with digital elevation model (DEM), land use type data, soil attribute data, meteorological data, and agricultural management data. In this way, the SWAT hydrological and water quality model is constructed and used to generate continuous long-series hydrological and pollutant load data. The flow monitoring data, pollutant monitoring data, and long-series hydrological and pollutant load data are integrated as hydrological and water quality monitoring data output.

[0022] Furthermore, the meteorological data includes: daily rainfall data, maximum temperature data, and minimum temperature data; the agricultural management data includes: crop type, planting time, harvest time, fertilizer type, fertilizer amount, fertilization time, and irrigation time.

[0023] S3. Establish the spatial response relationship of watershed water quality in the study area based on the target concentration. Specifically, this includes: determining the corresponding relationship between the main stream and tributary sections in the study area based on the target concentration, and testing the linear relationship between each section; determining whether the degree of linear relationship is greater than the relationship threshold. If so, the contribution coefficient method is executed, and the load contribution coefficient of each tributary to the watershed outlet is inferred by calculating the load ratio coefficient between adjacent sections; if not, the model prediction method is executed, and the nonlinear mapping relationship between the pollutant load of each tributary and the upstream section of the main stream and the pollutant load of the estuary section is predicted using a trained machine learning model to obtain the spatial response relationship of watershed water quality.

[0024] Furthermore, when the results output by the contribution coefficient method and the model prediction method do not meet the standard, the tributaries in the SWAT hydrological and water quality model are generalized as inputs.

[0025] Furthermore, the tributaries in the SWAT hydrological and water quality model are generalized as inputs, specifically including: constructing a SWAT hydrological and water quality model of the main stream area in the study area, and representing the tributary confluence points with Inlet points; the inputs of the Inlet points include: runoff and pollutant load.

[0026] Furthermore, based on the target concentration, the corresponding relationship between the main stream and tributary sections in the study area is determined, and the linear relationship between each section is tested. Specifically, starting from the upstream, if there is a tributary flowing into the first main stream section and the second main stream section, the target tributary is determined to be controlled by the second main stream section. The pollution load of the target tributary flowing into the river section intersects with the pollution load at the first main stream section, and the linear relationship between the pollution load of the target tributary and the second main stream section is fitted by the least squares method.

[0027] In practical applications, if the contribution coefficient method is used to establish the response relationship, the calculation formula for the total nitrogen load output at the watershed outlet is as follows: .

[0028] .

[0029] .

[0030] In the formula, For the first main stream from upstream to downstream One cross-section; This is the ratio of total nitrogen load between adjacent cross sections; For the first i Total nitrogen load at each main stream section; For the first The first cross section and the first The total nitrogen load of the main tributaries between each cross section; This refers to the number of cross-sections of the main stream. For the first The first cross section and the first The contribution coefficient of total nitrogen load of the main tributaries between cross sections.

[0031] The process of training a machine learning model requires data normalization; the training data only includes pollutant fluxes at each cross section; and cross-validation and hyperparameter optimization are performed during training.

[0032] S4. Taking the minimum difference in the percentage reduction of total nitrogen load in each tributary and the minimum total reduction of total nitrogen load in each tributary as the optimization objectives, and using the pollutant concentration or load control target at the estuary section as the constraint, a multi-objective optimization algorithm is constructed. Based on hydrological and water quality monitoring data and the spatial response relationship of watershed water quality, the multi-objective optimization algorithm iteratively optimizes the allowable pollutant discharge or reduction ratio at each control section to obtain the Pareto optimal solution set. Minimizing the total reduction of total nitrogen load in each tributary implies minimizing pollution control costs.

[0033] Furthermore, the objective function of the multi-objective optimization algorithm is expressed as follows: .

[0034] In the formula, The standard deviation of the percentage reduction in total nitrogen load between each cross section; This represents the total reduction amount at each cross-section; This represents the total nitrogen load reduction. This represents the original total nitrogen load.

[0035] Furthermore, the constraint expressions of the multi-objective optimization algorithm are as follows: .

[0036] The total nitrogen reduction at the sea section representing the control scheme is less than the control target; The total nitrogen reduction at the sea inlet section representing the control scheme is greater than 95% of the control target; The level of total nitrogen load control at the sea section when it meets the standards. The target for controlling total nitrogen load at the sea section.

[0037] In practical applications, the process of finding the Pareto optimal solution set is as follows: a multi-objective optimization algorithm is used to randomly generate pollutant load control schemes for each tributary; based on the contribution coefficient or a trained machine learning model, the water quality at the watershed outlet under each scheme is calculated, or the tributary control schemes are input into a generalized SWAT model to simulate the water quality at the watershed outlet; it is determined whether the water quality at the watershed outlet meets the objective; for schemes that meet the objective, the fairness (F1B) and cost reduction (F2B) of each scheme are calculated and then filtered; the above steps are repeated to finally obtain the Pareto optimal solution set.

[0038] The selection of the optimal solution includes: solutions that balance fairness (F1B) and cost reduction (F2B) and solutions that focus on fairness or cost reduction respectively.

[0039] S5. Analyze the characteristics of the Pareto frontier at different times in the Pareto optimal solution set, and based on the socio-economic and engineering feasibility of the watershed, select the control schemes for pollutants to be controlled that balance fairness and environmental benefits and efficiency from the Pareto optimal solution set, and determine the concentration control targets for pollutants to be controlled at each cross section.

[0040] As an optional implementation method, this embodiment takes total nitrogen as the pollutant to be controlled and a target concentration of 80% as an example to elaborate on the specific process of the pollution collaborative control optimization method based on watershed system.

[0041] Step 1, such as Figure 2 As shown, the pollutant to be controlled (the pollutant to be controlled) is identified as total nitrogen (TN), the study area is the Yellow River Basin, and the target concentration of this pollutant at the basin outlet Lijin is 80% of the target concentration in 2025.

[0042] Step 2, as follows Figure 3As shown, hydrological and water quality monitoring data for the study area were collected: control sections of the main stream and major tributaries of the basin were determined, and monthly-scale data on corresponding flow and pollutant monitoring were collected. For areas or periods with insufficient monitoring data, a SWAT model was constructed. The calibration sections were selected as Toudaoguai, Wubu, Longmen, Sanmenxia, ​​Huayuankou, and Lijin, and the tributary sections were selected as Baijiachuan (Wuding River, WDH), Hejin (Fen River, FH), Tongguan Suspension Bridge (Wei River, WH), Qilipu (Luo River, LH), and Wuzhi Canal Head (Qin River, QH). Flow and water quality data were collected. Monthly calibration and validation results showed that at the Lijin and Fen River sections, the coefficient of determination R² for runoff simulation was greater than 0.60, and the Nash efficiency coefficient NSE was greater than 0.50, meeting the basic requirements for hydrological simulation. The simulation effect was good for tributaries that significantly contributed to the total nitrogen load of the Yellow River (Fen River and Wei River sections), with R² > 0.70.

[0043] Step 3: As shown in Table 1, establish the spatial response relationship of watershed water quality: Based on the acquired data, test the linear relationship between each section. The specific method is as follows: Starting from the upstream, there are tributaries flowing between several flow sections 1 (first main stream section) and main stream section 2 (second main stream section). The tributary (target tributary) is controlled by main stream section 2. The pollution load of the tributary flowing into the river section needs to be intersected with the pollution load at main stream section 1. The relationship between the main stream section and the tributary section is fitted by machine learning XGBoost.

[0044] Table 1 Machine Learning Settings

[0045] The training results of the models were validated. The R2 of the vast majority of models was greater than 0.9. The validation results of the machine learning models are shown in Table 2.

[0046] Table 2 Validation Results of Machine Learning Models

[0047] Step 4: Construct a multi-objective optimization algorithm and generate control schemes: Set pollutant concentration or load control targets at the estuary section as constraints, and set multiple optimization objectives, including fairness of the reduction scheme and environmental efficiency. Use a multi-objective optimization algorithm to iteratively optimize the allowable pollutant emissions or reduction ratios at each control section to obtain the Pareto optimal solution set. The objective functions of the multi-objective optimization algorithm include the fairness objective F1B (minimizing the difference in the percentage reduction of total nitrogen load in each tributary) and the efficiency objective F2B (minimizing the total reduction). The specific process for obtaining the Pareto optimal solution set is as follows: The multi-objective optimization algorithm randomly generates pollutant load control schemes for each tributary; based on the machine learning model, the water quality at the watershed outlet under each scheme is calculated, and the Pareto solution set is as follows: Figure 4 As shown.

[0048] Step 5, as follows Figure 4 As shown, the optimal solution is selected based on the field scenario: The Pareto front characteristics at different hydrological periods are analyzed, and considering the socio-economic conditions of the watershed and the feasibility of engineering implementation, solutions that balance equity (F1B) and cost reduction (F2B) and solutions that focus on equity or cost reduction respectively are selected from the solution set. Pollutant concentration control targets for each cross section are determined, and the comprehensive optimal point (TB) is selected to analyze the corresponding total nitrogen concentration control target. For example... Figure 5 As shown, the results indicate that under non-flood season conditions, the average reduction rates of total nitrogen (TN) load under schemes F1B, F2B, and TB were 28.66%, 10.97%, and 21.06%, respectively, which were 8.86%, 1.02%, and 2.67% higher than those under flood season conditions. Among these, the Qinhe River (QH) had the highest reduction rate at 31.31%, followed by the Toudaoguai River (TDG) at 27.61%, both approximately twice the average reduction rates of other tributaries. The Weihe River (WH) had the lowest average reduction rate (11.53%), but still achieved a relatively high level (28.00%) under scheme F1B. The average reduction rate of TN load during the flood season was 4.18% lower than that during the non-flood season, with the Wuding River (Baijiachuan, WDH) showing the largest reduction at approximately 38.31%.

[0049] like Figure 6 As shown, under scenario F1B, with a 20% reduction target, the total nitrogen concentrations at the estuary sections during the non-flood season and flood season were 2.49 and 2.76 mg / L, respectively. Overall, the average reduction rate across all sections during the non-flood season reached 23.32%, 6.39% higher than during the flood season. Before reduction, the average total nitrogen concentration during the flood season was 0.36 mg / L lower than during the non-flood season; after reduction, this difference narrowed to 0.08 mg / L. Furthermore, there was a significant difference in total nitrogen removal efficiency between the main stream and tributaries. During the non-flood season, the average reduction rate of total nitrogen concentration in the main stream was 11.74% lower than that in the tributaries; while during the flood season, the removal rate in the main stream was still lower than that in the tributaries, but the gap narrowed to 6.32%. Especially during the non-flood season, the total nitrogen concentration in tributaries after reduction ranged from 1.03 to 4.74 mg / L, a decrease of 27.61% to 31.32% compared to the original concentration; while the total nitrogen concentration in the main stream after reduction ranged from 2.19 to 3.15 mg / L, an average decrease of 16.91% compared to the original concentration.

[0050] Example 2: This example provides a computer system, which can be a server or a terminal, and its internal structure diagram can be as follows. Figure 7As shown, the computer system includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media. The database stores forced oscillation samples and sub / supersynchronous oscillation samples. The I / O interfaces are used for information exchange between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements the aforementioned method for rapid prediction and identification of the dominant frequency of sub / supersynchronous oscillations in new energy power systems based on transfer learning.

[0051] Those skilled in the art will understand that Figure 7 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer system to which the present application is applied. A specific computer system may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0052] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0053] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0054] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0055] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A pollution collaborative control optimization method based on a watershed system, characterized in that, The method includes: Set the target concentration of the pollutant to be controlled at the watershed outlet within the study area; the study area is located within the target watershed system. Hydrological and water quality monitoring data of the study area were collected, and the default values ​​were supplemented by constructing and calibrating the SWAT hydrological and water quality model. The spatial response relationship of watershed water quality in the study area is established based on the target concentration. Specifically, this includes: determining the corresponding relationship between the main stream and tributary sections in the study area based on the target concentration, and testing the linear relationship between each section; determining whether the degree of linear relationship is greater than the relationship threshold. If so, the contribution coefficient method is executed, and the load contribution coefficient of each tributary to the watershed outlet is inferred by calculating the load ratio coefficient between adjacent sections; if not, the model prediction method is executed, and the nonlinear mapping relationship between the pollutant load of each tributary and the upstream section of the main stream and the pollutant load of the estuary section is predicted using a trained machine learning model to obtain the spatial response relationship of watershed water quality. The optimization objectives are to minimize the difference in the percentage reduction of total nitrogen load in each tributary and to minimize the total reduction of total nitrogen load in each tributary. The objectives are to use the pollutant concentration or load control target at the estuary section as the constraint. A multi-objective optimization algorithm is constructed, and based on hydrological and water quality monitoring data and the spatial response relationship of watershed water quality, the multi-objective optimization algorithm is used to iteratively optimize the allowable discharge or reduction ratio of pollutants at each control section to obtain the Pareto optimal solution set. The Pareto front characteristics at different periods in the Pareto optimal solution set were analyzed. Based on the socio-economic conditions of the watershed and the feasibility of engineering implementation, a control scheme for pollutants that balances fairness and environmental benefits and efficiency was selected from the Pareto optimal solution set, and the concentration control targets for pollutants at each cross section were determined.

2. The pollution collaborative control optimization method based on watershed systems according to claim 1, characterized in that, The target concentration can be measured on at least one of the following scales: monthly, seasonal, or annual.

3. The pollution collaborative control optimization method based on watershed systems according to claim 1, characterized in that, Hydrological and water quality monitoring data for the study area were collected, and the default values ​​were supplemented by constructing and calibrating a SWAT hydrological and water quality model. Specifically, this included: The control sections of the main stream and major tributaries of the watershed in the study area were determined, and corresponding flow monitoring data and pollutant monitoring data were collected. When there were areas or periods with insufficient monitoring data, digital elevation model data, land use type data, soil attribute data, meteorological data, and agricultural management data were collected instead. Based on this, the SWAT hydrological and water quality model was constructed and used to generate continuous long-series hydrological and pollutant load data. The flow monitoring data, pollutant monitoring data, and long-series hydrological and pollutant load data were integrated as the output hydrological and water quality monitoring data.

4. The pollution collaborative control optimization method based on watershed systems according to claim 3, characterized in that, The meteorological data includes: daily rainfall data, maximum temperature data, and minimum temperature data; The agricultural management data includes: crop type, planting time, harvest time, fertilizer type, fertilizer amount, fertilization time, and irrigation time.

5. The pollution collaborative control optimization method based on watershed systems according to claim 1, characterized in that, When the results output by the contribution coefficient method and the model prediction method do not meet the standard, the tributaries in the SWAT hydrological and water quality model are generalized as inputs.

6. The pollution collaborative control optimization method based on watershed systems according to claim 5, characterized in that, The tributaries in the SWAT hydrological and water quality model are generalized as inputs, specifically including: A SWAT hydrological and water quality model of the main stream area in the study region is constructed, and the inlet points are used to represent the tributary confluences. The inputs of the inlet points include runoff and pollutant load.

7. The pollution collaborative control optimization method based on watershed systems according to claim 1, characterized in that, The study determined the correspondence between the main stream and tributary sections in the study area based on the target concentration, and tested the linear relationship between each section. Specifically, starting from the upstream, if there is a tributary flowing into the first main stream section and the second main stream section, the target tributary is determined to be controlled by the second main stream section. The pollution load of the target tributary flowing into the river section intersects with the pollution load at the first main stream section, and the linear relationship between the pollution load of the target tributary and the second main stream section is fitted by the least squares method.

8. The pollution collaborative control optimization method based on watershed systems according to claim 1, characterized in that, The objective function of the multi-objective optimization algorithm is expressed as follows: ； In the formula, The standard deviation of the percentage reduction in total nitrogen load between each cross section; This represents the total reduction amount at each cross-section; This represents the total nitrogen load reduction. This represents the original total nitrogen load.

9. The pollution collaborative control optimization method based on watershed systems according to claim 1, characterized in that, The constraint expressions for the multi-objective optimization algorithm are as follows: ； The total nitrogen reduction at the sea section representing the control scheme is less than the control target; The total nitrogen reduction at the sea-entry section representing the control scheme is greater than 95% of the control target; The level of total nitrogen load control at the sea section when it meets the standards. The target for controlling total nitrogen load at the sea section.

10. A computer system, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the pollution collaborative control optimization method based on any one of claims 1-9.

Images