Water ecological pollution comprehensive treatment method

Abstract

The present application relates to a kind of water ecological pollution comprehensive management method, comprising the following specific steps: collecting water quality parameter, hydrological parameter, pollution source parameter and meteorological parameter, constructs multidimensional original data set, carries out data cleaning, abnormal value rejection, data standardization operation to water quality parameter, hydrological parameter, pollution source parameter and meteorological parameter, obtains standardization data set, establishes water quality prediction model and pollution traceability model based on standardization data set, combines water quality prediction result and pollution source contribution degree, matches the historical management scheme of historical management database, obtains optimal management scheme, optimization is carried out based on the management effect of management scheme.The water ecological pollution comprehensive management method, not only can collect and analyze various factors influencing water quality, can accurately predict water quality change and find pollution source, and dynamically adjust management strategy according to real-time data and historical experience, save management resource, improve management effect.

Description

Technical Field

[0001] This invention relates to the field of water environment management technology, specifically to a comprehensive method for the management of water ecological pollution. Background Technology

[0002] Water pollution control treats the water system as a whole, taking into account the geographical distribution of towns and industrial and mining enterprises along the water system, as well as the water system's self-purification capacity, pollution carrying capacity, and pollution status. It is a comprehensive set of prevention and control measures for preventing and controlling water pollution. It includes various engineering and technical means and management measures, and has the characteristics of being holistic, comprehensive, and regional. Its core objectives include pollutant emission reduction, water ecological restoration, and the construction of a long-term management mechanism.

[0003] Chinese invention patent authorization announcement number CN117455739B proposed an intelligent water environment comprehensive management method and system. However, it only collected data on single factors such as water quality or water flow, without comprehensively considering multiple factors such as pollution sources and meteorological conditions. This resulted in inaccurate prediction of water pollution, difficulty in accurately locating the source of pollution, and that the management plan was often relatively fixed and could not be flexibly adjusted according to the actual situation, thus wasting management resources and affecting the management effect. Therefore, an improvement is needed. Hence, a comprehensive water ecological pollution management method is proposed to solve the above problems. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a comprehensive water ecological pollution control method with advantages such as good control effect. It solves the problems of only collecting data on single factors such as water quality or water flow without comprehensively considering multiple factors such as pollution sources and meteorological conditions, which leads to inaccurate prediction of water pollution, difficulty in accurately locating the source of pollution, and that the control plan is often relatively fixed and cannot be flexibly adjusted according to the actual situation, thus wasting control resources and affecting the control effect.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a comprehensive method for water ecological pollution control, comprising the following specific steps: S1. Collect water quality parameters, hydrological parameters, pollution source parameters, and meteorological parameters to construct a multi-dimensional raw dataset. ,in For the sample size, As an indicator dimension; S2. Perform data cleaning, outlier removal, and data standardization on water quality parameters, hydrological parameters, pollution source parameters, and meteorological parameters to obtain a standardized dataset; S3. Establish water quality prediction models and pollution source tracing models based on standardized datasets; S3.1. Construct a water quality prediction model based on the LSTM model and the random forest model, and obtain the water quality prediction results; S3.2. Construct a pollution source tracing model based on a Bayesian network model to obtain the contribution of pollution sources; S4. Combining water quality prediction results with pollution source contribution, matching historical treatment schemes from the historical treatment database, and obtaining the optimal treatment scheme; S4.1. Calculate the dosage of the reagent based on the turbidity of the water and the degradation rate constant of the pollutants; S4.2. Based on the water turbidity and pollutant degradation rate constant, adopt the corresponding reagent dosing scheme; S4.3. Calculate the area of ​​constructed wetlands based on the average daily treated water volume, influent pollutant concentration, and effluent pollutant concentration.

[0006] S5. Optimize based on the governance effectiveness of the governance scheme.

[0007] Furthermore, the water quality parameters mentioned in step S1 include pH value, dissolved oxygen, ammonia nitrogen, total phosphorus, and turbidity; the hydrological parameters include water level, flow velocity, and flow rate; the pollution source parameters include discharge concentration, discharge volume, and discharge outlet location; and the meteorological parameters include rainfall, air temperature, and wind speed.

[0008] Furthermore, the data cleaning in step S2 uses the weighted mean imputation method to handle missing values, and the calculation formula is as follows: In the formula, For the data after filling in, This is the original missing data. This represents the number of non-missing samples for this indicator. For the first The first non-missing sample Item index value, Weighting coefficients By collection time interval Assignment; Outlier removal uses a modified Z-score method, calculated as follows: In the formula, For the first The median of the indicators 4826 is the absolute median difference, and 4826 is the correction coefficient for a normal distribution. At that time, the data was considered an outlier and was removed. Data standardization uses Z-score standardization to eliminate the influence of dimensions. The calculation formula is as follows: In the formula, For standardized data, For the first The mean of the indicators, For the first The standard deviation of each indicator.

[0009] Furthermore, in step S3.1, the number of hidden layer neurons in the LSTM model is 64-128, the number of decision trees in the random forest model is 100-200, and the calculation formula for the water quality prediction model is: In the formula, for Predicted water quality values ​​at any time The fusion weights are set between 0.6 and 0.8, and the optimal values ​​are determined through cross-validation. for The input feature vector contains preprocessed water quality, hydrological, and meteorological data. Output results for the LSTM model. This is the output of the random forest model.

[0010] Furthermore, the nodes of the Bayesian network described in step S3.2 include pollution source type, pollutant concentration, and water quality exceedance status. The pollution source type is divided into industrial pollution source, domestic pollution source, and agricultural non-point source. The conditional probability table is trained based on historical monitoring data. The calculation formula for the pollution source tracing model is as follows: In the formula, For the first Individual pollution sources caused water quality exceeding standards. The posterior probability, i.e., the contribution. As a source of pollution The conditional probability of water quality exceeding standards when it occurs. As a source of pollution The prior probability is obtained based on statistical analysis of historical emissions data. This represents the total number of pollution sources.

[0011] Furthermore, the reagents mentioned in step S4.1 include polyaluminum chloride and polyacrylamide, and the formula for calculating the dosage of the reagents is as follows: In the formula, This refers to the dosage of the drug. To manage the volume of water, This represents the current pollutant concentration. For the target pollutant concentration, For safety factor, safety factor At low temperature ( ) or high turbidity ( In a given environment, the value is 3; in a normal environment, the value is 1. For reagent purification efficiency, the corresponding polyaluminum chloride The value is 85%-90%, corresponding to polyacrylamide. The value is 90%-95%. This is the turbidity correction factor, when hour, ,when hour, ,when hour, , Dissolved oxygen concentration in water. For water temperature, is the pollutant degradation rate constant.

[0012] Furthermore, the agent dosing scheme described in step S4.2 includes the following specific steps: S4.2.1. When and At the time of application, the initial dosage is 100% of the value calculated by the formula. The pollutant concentration is monitored every 2 hours. If the concentration decrease rate is ≥0.1mg / (L·h), the current dosage is maintained. If the concentration decrease rate is <0.1mg / (L·h), 10% of the agent calculated by the formula is added. S4.2.2. When and The agent is added in three batches: 60% of the calculated value is added for the first batch, 30% is added after 4 hours, and the remaining 10% is added after 8 hours based on the concentration monitoring results. At the same time, aeration equipment is used to improve the water flow. S4.2.3. When and First, add a coagulant aid at a rate of 10%-15% of the main agent to reduce turbidity to below 60 NTU. Then, recalculate the agent dosage according to the above formula and apply it to the upstream, midstream, and downstream areas of the watershed. The dosage in the core pollution area of ​​the midstream watershed should be increased by 20% compared to the calculated value, while the dosage in the upstream and downstream areas should be based on the calculated value.

[0013] Furthermore, the formula for calculating the area of ​​the constructed wetland mentioned in step S4.3 is as follows: In the formula, Area of ​​artificial wetlands This represents the average daily water treatment volume. The concentration of pollutants in the influent. The concentration of pollutants in the effluent. Let be the pollutant removal rate constant. For ammonia nitrogen, the concentration is 0.05-0.08 mg / d, and for total phosphorus, it is 0.03-0.06 mg / d. The depth of the wetland water. The depth is 0.6-1.0m. Hydraulic residence time It takes 2-5 days.

[0014] Furthermore, the optimization of governance effectiveness based on the governance scheme described in step S5 includes the following specific steps: S5.1. Calculate the compliance rate of the treatment effect. The calculation formula is as follows: In the formula, To achieve the target rate, To meet the required number of monitoring sessions, Total number of monitoring sessions; S5.2. Feed the treated data back to the water quality prediction model, and update the model parameters using the gradient descent method. The calculation formula is as follows: In the formula, For the updated parameters, For parameters before the update, For learning rate, The value range is 0.001-0.01. This represents the gradient of the loss function with the current parameters. S5.3. Based on the model iteration results and the compliance rate, dynamically adjust the dosage of the agent, the dosage scheme, and the ecological restoration scale parameters to optimize the process.

[0015] Furthermore, the loss function described in step S5.2 uses mean squared error, and the calculation formula is as follows: In the formula, This represents the actual water quality value. To predict water quality values, the model iteration cycle is 72 hours.

[0016] Compared with the prior art, the technical solution of this application has the following beneficial effects: This comprehensive water pollution control method can not only collect and analyze various factors affecting water quality, accurately predict water quality changes and locate pollution sources, but also dynamically adjust control strategies based on real-time data and historical experience, saving control resources and improving control effectiveness. Attached Figure Description

[0017] Figure 1 This is a flowchart of the present invention; Figure 2 This is a sub-flowchart of the optimal governance scheme of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Please see Figure 1-2 The water ecological pollution comprehensive treatment method in this embodiment includes the following specific steps: S1. Collect water quality parameters, hydrological parameters, pollution source parameters, and meteorological parameters. Water quality parameters include pH, dissolved oxygen, ammonia nitrogen, total phosphorus, and turbidity. Hydrological parameters include water level, flow velocity, and flow rate. Pollution source parameters include discharge concentration, discharge volume, and discharge outlet location. Meteorological parameters include rainfall, temperature, and wind speed. Construct a multi-dimensional raw dataset. ,in For the sample size, As an indicator dimension; S2. Perform data cleaning, outlier removal, and data standardization on water quality parameters, hydrological parameters, pollution source parameters, and meteorological parameters to obtain a standardized dataset; Data cleaning uses the weighted mean imputation method to handle missing values. The calculation formula is as follows: In the formula, For the data after filling in, This is the original missing data. This represents the number of non-missing samples for this indicator. For the first The first non-missing sample Item index value, Weighting coefficients By collection time interval Assignment, when hour, ;when hour, ;when hour, ; Outlier removal uses a modified Z-score method, calculated as follows: In the formula, For the first The median of the indicators 4826 is the absolute median difference, and 4826 is the correction coefficient for a normal distribution. At that time, the data was considered an outlier and was removed. Data standardization uses Z-score standardization to eliminate the influence of dimensions. The calculation formula is as follows: In the formula, For standardized data, For the first The mean of the indicators, For the first The standard deviation of the indicators; S3. Establish water quality prediction models and pollution source tracing models based on standardized datasets; S3.1. Construct a water quality prediction model based on the LSTM model and the random forest model, and obtain the water quality prediction results; The LSTM model has 64-128 hidden layer neurons, the random forest model has 100-200 decision trees, and the calculation formula for the water quality prediction model is: In the formula, for Predicted water quality values ​​at any time The fusion weights are set between 0.6 and 0.8, and the optimal values ​​are determined through cross-validation. for The input feature vector contains preprocessed water quality, hydrological, and meteorological data. Output results for the LSTM model. Output results for the random forest model; S3.2. Construct a pollution source tracing model based on a Bayesian network model to obtain the contribution of pollution sources; The nodes of the Bayesian network include pollution source type, pollutant concentration, and water quality exceedance status. Pollution source types are categorized into industrial, domestic, and agricultural non-point source pollution. The conditional probability table is trained based on historical monitoring data. The calculation formula for the pollution source tracing model is as follows: In the formula, For the first Individual pollution sources caused water quality exceeding standards. The posterior probability, i.e., the contribution. As a source of pollution The conditional probability of water quality exceeding standards when it occurs. As a source of pollution The prior probability is obtained based on statistical analysis of historical emissions data. The total number of pollution sources; S4. Combining water quality prediction results with pollution source contribution, matching historical treatment schemes from the historical treatment database, and obtaining the optimal treatment scheme; S4.1. Calculate the dosage of the reagent based on the turbidity of the water and the degradation rate constant of the pollutants; The reagents include polyaluminum chloride and polyacrylamide. The formula for calculating the dosage of the reagents is as follows: In the formula, This refers to the dosage of the drug. To manage the volume of water, This represents the current pollutant concentration. For the target pollutant concentration, For safety factor, safety factor At low temperature ( ) or high turbidity ( In a given environment, the value is 3; in a normal environment, the value is 1. For reagent purification efficiency, the corresponding polyaluminum chloride The value is 85%-90%, corresponding to polyacrylamide. The value is 90%-95%. This is the turbidity correction factor, when hour, ,when hour, ,when hour, , Dissolved oxygen concentration in water. For water temperature, This represents the pollutant degradation rate constant. S4.2. Based on the water turbidity and pollutant degradation rate constant, adopt the corresponding reagent dosing scheme, including the following specific steps: S4.2.1. When and At the time of application, the initial dosage is 100% of the value calculated by the formula. The pollutant concentration is monitored every 2 hours. If the concentration decrease rate is ≥0.1mg / (L·h), the current dosage is maintained. If the concentration decrease rate is <0.1mg / (L·h), 10% of the agent calculated by the formula is added. S4.2.2. When and The agent is added in three batches: 60% of the calculated value is added for the first batch, 30% is added after 4 hours, and the remaining 10% is added after 8 hours based on the concentration monitoring results. At the same time, aeration equipment is used to improve the water flow. S4.2.3. When and First, add a coagulant aid at a rate of 10%-15% of the main agent to reduce turbidity to below 60 NTU. Then, recalculate the agent dosage according to the above formula and apply it to the upstream, midstream, and downstream areas of the watershed. The dosage in the core pollution area of ​​the midstream watershed should be increased by 20% compared to the calculated value, while the dosage in the upstream and downstream areas should be based on the calculated value. S4.3. Calculate the area of ​​constructed wetlands based on the average daily treated water volume, influent pollutant concentration, and effluent pollutant concentration; The formula for calculating the area of ​​constructed wetlands is: In the formula, Area of ​​artificial wetlands This represents the average daily water treatment volume. The concentration of pollutants in the influent. The concentration of pollutants in the effluent. Let be the pollutant removal rate constant. For ammonia nitrogen, the concentration is 0.05-0.08 mg / d, and for total phosphorus, it is 0.03-0.06 mg / d. The depth of the wetland water. The depth is 0.6-1.0m. Hydraulic residence time For 2-5 days; S5. Optimize based on the governance effectiveness of the governance solution, including the following specific steps: S5.1. Calculate the compliance rate of the treatment effect. The calculation formula is as follows: In the formula, To achieve the target rate, To meet the required number of monitoring sessions, Total number of monitoring sessions; S5.2. Feed the treated data back to the water quality prediction model, and update the model parameters using the gradient descent method. The calculation formula is as follows: In the formula, For the updated parameters, For parameters before the update, For learning rate, The value range is 0.001-0.01. This represents the gradient of the loss function with the current parameters. The loss function uses mean squared error, and the calculation formula is as follows: In the formula, This represents the actual water quality value. To predict water quality values, the model iteration period is 72 hours; S5.3. Based on the model iteration results and the compliance rate, dynamically adjust the dosage of the agent, the dosage scheme, and the ecological restoration scale parameters to optimize the process.

[0020] As a specific example: Data was collected continuously for 30 days, including water quality parameters such as pH (7.2-8.5), dissolved oxygen (4.5-6.8 mg / L), ammonia nitrogen (0.3-0.8 mg / L), total phosphorus (0.1-0.3 mg / L), and turbidity (25-70 NTU); hydrological parameters such as water level (2.5-3.2 m) and flow velocity (0.3-0.6 m / s); pollution source parameters such as discharge concentration and discharge volume from three main sewage outlets; and meteorological parameters such as rainfall (5-20 mm / d) and temperature (18-28℃). This data was used to construct the original dataset. ; The degradation rate calculation submodule 17 calculates the degradation rate according to the formula. Calculate the pollutant degradation rate constant at a certain monitoring point. , ,but ; After data cleaning, the weighted mean imputation method was used to process 120 missing data points, including the missing ammonia nitrogen values ​​at a certain monitoring point. In the non-missing samples of this indicator, there are 3 recent ( )sample , , ( ), 2 long-term ( )sample , ( ),but , After filling ; By removing outliers from the data, the total phosphorus data for a certain monitoring point was calculated. , A certain data ,but These are identified as outliers and removed. Ammonia nitrogen index was calculated by standardizing the data. , A certain data After standardization ; By setting Input the standardized feature vector The fusion model was trained to predict ammonia nitrogen concentrations for the next 7 days, with a prediction error of 3.2%. The contribution of three pollution sources was calculated: industrial wastewater outlets Domestic sewage outlet Agricultural non-point source The core risk point is identified as industrial wastewater discharge outlets; PAC agent (selected) ), Treatment of water volume , , , a certain area ( ), ,but ; because and Choose the third dosing option, first add the coagulant aid ( Two hours later, the turbidity dropped to 55 NTU, and the readings were recalculated. , Distribution according to watershed zones: midstream addition Upstream and downstream companies each add ; Constructing artificial wetlands ammonia nitrogen , , , , ,but ; After treatment, continuous monitoring for 7 days showed that the average ammonia nitrogen concentration decreased to 0.18 mg / L. , compliance rate ; use After updating the model parameters and iterating, the prediction error decreased to 2.8%. Subsequent monitoring revealed Upgraded to , Once the concentration is stabilized at 40 NTU, the second dosing option is selected, increasing the dosage to 42,000 kg to ensure both precision and cost-effectiveness in treatment.

[0021] The working principle of the above embodiments is as follows: This comprehensive water pollution control method can not only collect and analyze various factors affecting water quality, accurately predict water quality changes and locate pollution sources, but also dynamically adjust control strategies based on real-time data and historical experience, saving control resources and improving control effectiveness.

[0022] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0023] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A comprehensive method for water ecological pollution control, characterized in that, The specific steps include the following: S1. Collect water quality parameters, hydrological parameters, pollution source parameters, and meteorological parameters to construct a multi-dimensional raw dataset. ,in For the sample size, As an indicator dimension; S2. Perform data cleaning, outlier removal, and data standardization on water quality parameters, hydrological parameters, pollution source parameters, and meteorological parameters to obtain a standardized dataset; S3. Establish water quality prediction models and pollution source tracing models based on standardized datasets; S3.

1. Construct a water quality prediction model based on the LSTM model and the random forest model, and obtain the water quality prediction results; S3.

2. Construct a pollution source tracing model based on a Bayesian network model to obtain the contribution of pollution sources; S4. Combining water quality prediction results with pollution source contribution, matching historical treatment schemes from the historical treatment database, and obtaining the optimal treatment scheme; S4.

1. Calculate the dosage of the reagent based on the turbidity of the water and the degradation rate constant of the pollutants; S4.

2. Based on the water turbidity and pollutant degradation rate constant, adopt the corresponding reagent dosing scheme; S4.

3. Calculate the area of ​​constructed wetlands based on the average daily treated water volume, influent pollutant concentration, and effluent pollutant concentration. S5. Optimize based on the governance effectiveness of the governance scheme.

2. The comprehensive water ecological pollution control method according to claim 1, characterized in that, The water quality parameters mentioned in step S1 include pH value, dissolved oxygen, ammonia nitrogen, total phosphorus, and turbidity; the hydrological parameters include water level, flow velocity, and flow rate; the pollution source parameters include discharge concentration, discharge volume, and discharge outlet location; and the meteorological parameters include rainfall, air temperature, and wind speed.

3. The comprehensive water ecological pollution control method according to claim 1, characterized in that, The data cleaning in step S2 uses the weighted mean imputation method to handle missing values. The calculation formula is as follows: In the formula, For the data after filling in, This is the original missing data. This represents the number of non-missing samples for this indicator. For the first The j-th index value of non-missing samples, Weighting coefficients By collection time interval Assignment; Outlier removal uses a modified Z-score method, calculated as follows: In the formula, For the first The median of the indicators 4826 is the absolute median difference, and 4826 is the correction coefficient for a normal distribution. At that time, the data was considered an outlier and was removed. Data standardization uses Z-score standardization to eliminate the influence of dimensions. The calculation formula is as follows: In the formula, For standardized data, For the first The mean of the indicators, For the first The standard deviation of each indicator.

4. The comprehensive water ecological pollution control method according to claim 1, characterized in that, In step S3.1, the number of hidden layer neurons in the LSTM model is 64-128, the number of decision trees in the random forest model is 100-200, and the calculation formula for the water quality prediction model is: In the formula, for Predicted water quality values ​​at any time The fusion weights are set between 0.6 and 0.8, and the optimal values ​​are determined through cross-validation. for The input feature vector contains preprocessed water quality, hydrological, and meteorological data. Output results for the LSTM model. This is the output of the random forest model.

5. The comprehensive water ecological pollution control method according to claim 4, characterized in that, Step S3.2 The nodes of the Bayesian network include pollution source type, pollutant concentration, and water quality exceedance status. The pollution source types are divided into industrial pollution sources, domestic pollution sources, and agricultural non-point source pollution sources. The conditional probability table is trained based on historical monitoring data. The calculation formula for the pollution source tracing model is: In the formula, For the first Individual pollution sources caused water quality exceeding standards. The posterior probability, i.e., the contribution. As a source of pollution The conditional probability of water quality exceeding standards when it occurs. As a source of pollution The prior probability is obtained based on statistical analysis of historical emissions data. This represents the total number of pollution sources.

6. The comprehensive water ecological pollution control method according to claim 1, characterized in that, The reagents mentioned in step S4.1 include polyaluminum chloride and polyacrylamide. The formula for calculating the dosage of the reagents is as follows: In the formula, This refers to the dosage of the drug. To manage the volume of water, This represents the current pollutant concentration. For the target pollutant concentration, For safety factor, safety factor At low temperature ( ) or high turbidity ( In a given environment, the value is 3; in a normal environment, the value is 1. For reagent purification efficiency, the corresponding polyaluminum chloride The value is 85%-90%, corresponding to polyacrylamide. The value is 90%-95%. This is the turbidity correction factor, when hour, ,when hour, ,when hour, , Dissolved oxygen concentration in water. For water temperature, is the pollutant degradation rate constant.

7. The comprehensive water ecological pollution control method according to claim 1, characterized in that, Step S4.2 The reagent dosing scheme includes the following specific steps: S4.2.

1. When and At the time of application, the initial dosage is 100% of the value calculated by the formula. The pollutant concentration is monitored every 2 hours. If the concentration decrease rate is ≥0.1mg / (L·h), the current dosage is maintained. If the concentration decrease rate is <0.1mg / (L·h), 10% of the agent calculated by the formula is added. S4.2.

2. When and The agent is added in three batches: 60% of the calculated value is added for the first batch, 30% is added after 4 hours, and the remaining 10% is added after 8 hours based on the concentration monitoring results. At the same time, aeration equipment is used to improve the water flow. S4.2.

3. When and First, add a coagulant aid at a rate of 10%-15% of the main agent to reduce turbidity to below 60 NTU. Then, recalculate the agent dosage according to the above formula and apply it to the upstream, midstream, and downstream areas of the watershed. The dosage in the core pollution area of ​​the midstream watershed should be increased by 20% compared to the calculated value, while the dosage in the upstream and downstream areas should be based on the calculated value.

8. The comprehensive water ecological pollution control method according to claim 1, characterized in that, The formula for calculating the area of ​​the constructed wetland mentioned in step S4.3 is as follows: In the formula, Area of ​​artificial wetlands This represents the average daily water treatment volume. The concentration of pollutants in the influent. The concentration of pollutants in the effluent. Let be the pollutant removal rate constant. For ammonia nitrogen, the concentration is 0.05-0.08 m³ / d, and for total phosphorus, it is 0.03-0.06 m³ / d. The depth of the wetland water. The depth is 0.6-1.0m. Hydraulic residence time It takes 2-5 days.

9. The comprehensive water ecological pollution control method according to claim 1, characterized in that, Step S5, which optimizes the governance effect based on the governance scheme, includes the following specific steps: S5.

1. Calculate the compliance rate of the treatment effect. The calculation formula is as follows: In the formula, To achieve the target rate, To meet the required number of monitoring sessions, Total number of monitoring sessions; S5.

2. Feed the treated data back to the water quality prediction model, and update the model parameters using the gradient descent method. The calculation formula is as follows: In the formula, For the updated parameters, For parameters before the update, For learning rate, The value range is 0.001-0.

01. This represents the gradient of the loss function with the current parameters. S5.

3. Based on the model iteration results and the compliance rate, dynamically adjust the dosage of the agent, the dosage scheme, and the ecological restoration scale parameters to optimize the process.

10. The comprehensive water ecological pollution control method according to claim 9, characterized in that, The loss function described in step S5.2 uses mean squared error, and the calculation formula is as follows: In the formula, This represents the actual water quality value. To predict water quality values, the model iteration cycle is 72 hours.

Images