An overflow sewage coagulation-oxidation treatment method based on support vector regression and particle swarm optimization algorithm

Abstract

The present invention discloses a coagulation-oxidation treatment method for overflow sewage based on support vector regression and particle swarm optimization algorithm in the technical field of sewage treatment methods, including selecting the optimal treatment method for overflow sewage from various sewage treatment chemical addition methods. Under the conditions of the optimal treatment method, the effluent pollutant concentrations under the influence of individual characteristic variables are measured respectively, and a prediction model of the water quality index (WQI) of the effluent is constructed based on the pollutant concentrations. By comprehensively considering the co-removal of various pollutants and using the support vector regression (SVR) and particle swarm optimization algorithm (PSO) to optimize the coupling process parameters of sodium hypochlorite (NaClO) oxidation and polyaluminum chloride (PACl) coagulation, the present invention can effectively improve the treatment effect of urban overflow sewage and ensure water environment safety and sustainable development.

Description

Technical Field

[0001] The present invention relates to the technical field of sewage treatment methods, and specifically, it is a coagulation-oxidation treatment method for overflow sewage based on support vector regression and particle swarm optimization algorithms. Background Technique

[0002] Currently, there is a common problem of rain-sewage mixed connection in urban drainage pipes, which leads to an increased load on the drainage pipe network during heavy rain or continuous rainfall, and is prone to the problem of rain-sewage overflow. Overflow sewage has pollution characteristics such as multi-source, multi-state, and impact. If pollutants such as high-concentration ammonia nitrogen (NH3-N), chemical oxygen demand (COD), total nitrogen (TN), organic matter, total phosphorus (TP), and suspended solids (SS) are directly discharged without treatment, it will seriously pollute the receiving water body and threaten the balance of the water ecological system and the safety of residents' drinking water.

[0003] Effectively controlling overflow pollution has become an urgent need to improve the sustainable development of cities, and developing effective end-treatment technologies is crucial for improving the quality of the urban water environment.

[0004] The Chinese invention patent with the publication number CN119294612A provides a sewage treatment plant parameter prediction method and system based on machine learning, which shows significant advantages in the prediction of key operating parameters of sewage treatment plants. However, the impact pollution characteristics of overflow sewage make it difficult to directly apply the processes applicable in sewage treatment plants to the treatment of overflow sewage. In addition, the research on machine learning algorithms for optimizing the process parameters of overflow treatment is still relatively limited, and it is difficult to effectively improve the efficiency and water quality of sewage treatment. Summary of the Invention

[0005] The purpose of the present invention is to provide a coagulation-oxidation treatment method for overflow sewage based on support vector regression and particle swarm optimization algorithms. By comprehensively considering the co-removal of multiple pollutants and using support vector regression (SVR) and particle swarm optimization algorithms (PSO) to optimize the coupling process parameters of sodium hypochlorite (NaClO) oxidation and polyaluminum chloride (PACl) coagulation, the treatment effect of urban overflow sewage can be effectively improved.

[0006] To achieve the above purpose, the present invention provides the following technical solutions:

[0007] A coagulation-oxidation treatment method for overflow sewage based on support vector regression and particle swarm optimization algorithms, the method is used to control the concentration of pollutants in overflow sewage, and the method includes:

[0008] Select the optimal treatment method for overflow sewage from various ways of adding sewage treatment agents, which can achieve the pretreatment of overflow sewage. By performing a correlation analysis on various characteristic variables affecting the effluent quality and using a standardization method to eliminate the dimensional difference, the convergence speed of the model can be improved;

[0009] Under the conditions of the optimal treatment method, measure the effluent pollutant concentration under the influence of a single characteristic variable respectively, and judge the treatment effect according to the correlation between the pollutant concentration and the water quality cleanliness;

[0010] Build a prediction model of the effluent water quality index (WQI) based on the pollutant concentration.

[0011] As a further solution of the present invention: Preferably, the pollutants are SS, COD, TN, NH3-N and TP. Take the overflow sewage from the sewage pipeline, and its water quality characteristics are: SS = 396 mg / L, NH3-N = 21.14 mg / L, TN = 22.52 mg / L, COD = 302.5 mg / L, TP = 1.32 mg / L, pH = 8.1 and water temperature = 25 °C.

[0012] As a further solution of the present invention: The method for selecting the optimal treatment method for overflow sewage from various ways of adding sewage treatment agents includes:

[0013] Take the overflow sewage to be treated in a reaction container, add sewage treatment agents to the reaction container in a certain order to make the overflow sewage stratified;

[0014] Calculate the water quality pollution index (WQI) of the overflow sewage to obtain the result data of the overflow sewage;

[0015] Compare all the result data pairwise, select the result data corresponding to the smallest water quality pollution index (WQI) as the optimal result data, and use the treatment method that obtains the optimal result data as the optimal treatment method. By comparing and analyzing the influence of different chemical agent addition sequences on the treatment results of overflow sewage, the optimal chemical agent addition sequence and the treatment method of the chemical agent can be quickly judged, which is beneficial to improving the water treatment effect and treatment efficiency of overflow sewage.

[0016] As a further solution of the present invention: The first water treatment agent is a coagulant. By using a coagulant as the first water treatment agent, the tiny particles in the sewage can be aggregated into flocs through chemical or physical actions, which is convenient for subsequent sedimentation and filtration treatment. The second water treatment agent is an oxidant, and the oxidant can oxidize the COD in the sewage, which is beneficial to removing the organic substances in the sewage through oxidation, improving the odor in the sewage, and reducing the chromaticity of the sewage through the oxidation of the organic pigments in the sewage.

[0017] As a further solution of the present invention: the coagulant is a polyaluminium chloride (PACl) solution, the concentration of the polyaluminium chloride (PACl) solution is 250 mg / L, and the oxidant is a sodium hypochlorite (NaClO) solution, the concentration of the sodium hypochlorite (NaClO) solution is 200 mg / L.

[0018] As a further solution of the present invention: the method for calculating the water quality index (WQI) of overflow sewage to obtain result data of overflow sewage includes:

[0019] Using the pollution index calculation formula Calculate the water pollution index, where i is the type of pollutant; I i is the water pollution index of pollutant i; C i is the concentration of pollutant i; S i The evaluation standard for pollutant i is based on the three-level standard in the national standard "Pollutant Discharge Standard for Urban Wastewater Treatment Plants" (GB 18918-2002) currently in force in water bodies. The smaller the WQI value, the cleaner the water quality.

[0020] As a further embodiment of the present invention, the method of adding sewage treatment agents to the reaction vessel in a certain order comprises:

[0021] Adding a first water treatment agent and a second water treatment agent simultaneously to a first reaction container to obtain a first solution; forming a simultaneous coagulation-oxidation treatment method;

[0022] The method for stratifying overflow sewage comprises:

[0023] After stirring the first solution at a speed of 250 r / min for 30 seconds, the first solution was stirred at a speed of 50 r / min for 4 to 5 minutes;

[0024] The stirred first solution was allowed to settle for 5 minutes to obtain a first layered solution;

[0025] The method for calculating the water quality index (WQI) of overflow sewage to obtain the result data of overflow sewage includes:

[0026] measuring the pollutant concentration value in the first layer solution;

[0027] Based on the pollutant concentration, the water quality index (WQI) of the first layered solution is calculated using the pollution index calculation formula, and a first pollution index is obtained according to the calculation result.

[0028] As a further embodiment of the present invention, the method of adding sewage treatment agents to the reaction vessel in a certain order to stratify the overflow sewage further comprises:

[0029] Add the first water treatment agent to the second reaction vessel to obtain a second solution, constituting a pre-coagulation-oxidation treatment method. Stir the second solution at a speed of 250 r / min for 1 min, and after the stirring ends, let the second solution stand for 5 min;

[0030] Add the second water treatment agent to the second solution after standing. Stir the second solution with the second water treatment agent added at a speed of 100 r / min for 5 min, and after the stirring ends, let the second solution stand and precipitate for 5 min to obtain a second stratified solution;

[0031] The method for calculating the water quality pollution index (WQI) of the overflow sewage to obtain the result data of the overflow sewage further includes:

[0032] Measure the pollutant concentration in the second stratified solution;

[0033] Based on the pollutant concentration, use the pollution index calculation formula to calculate the water quality pollution index (WQI) of the second stratified solution, and obtain the second pollution index according to the calculation result.

[0034] As a further solution of the present invention: The method for adding the sewage treatment agent to the reaction vessel in a certain order to stratify the overflow sewage further includes:

[0035] Add the second water treatment agent to the third reaction vessel to obtain a third solution. Stir the third solution at a speed of 50 r / min for 5 min, and add the first water treatment agent to the stirred third solution; constituting a pre-oxidation-coagulation treatment method;

[0036] After stirring at a speed of 250 r / min for 1 min, let it stand and precipitate for 5 min to obtain a third stratified solution;

[0037] The method for calculating the water quality pollution index (WQI) of the overflow sewage to obtain the result data of the overflow sewage further includes:

[0038] Measure the pollutant concentration in the third stratified solution;

[0039] Based on the pollutant concentration, use the pollution index calculation formula to calculate the water quality pollution index (WQI) of the third stratified solution, and obtain the third pollution index according to the calculation result.

[0040] As a further solution of the present invention: The single characteristic variable includes any one of the factors of coagulant concentration value, oxidant concentration value, coagulation time, oxidation time, pH value, water temperature value, and influent water quality data value. By introducing characteristic variables including pH value, water temperature value, and influent water quality, it is beneficial to improve the prediction accuracy of the model by increasing the breadth of the characteristic variables;

[0041] The method for measuring the concentration of effluent pollutants under the influence of a single characteristic variable under the optimal treatment method includes:

[0042] According to the optimal treatment method, each time the value of one factor in the characteristic variable is changed, the concentration of pollutants in the effluent is measured to obtain the pollutant concentration value.

[0043] The method for constructing a prediction model of the effluent water quality index (WQI) based on the pollutant concentration includes:

[0044] Use the support vector regression (SVR) model to predict the effluent water quality index, where the effluent water quality index is the pollutant concentration, and use the particle swarm optimization algorithm (PSO) to optimize the hyperparameters of the support vector regressor (SVR) model. The hyperparameters include the penalty factor C and the kernel function parameter gamma, and construct a PSO-SVR effluent water quality prediction model.

[0045] As a further solution of the present invention: The method for optimizing the hyperparameters of the support vector regressor (SVR) model using the particle swarm optimization algorithm (PSO) includes:

[0046] Set the target parameter range. The parameters include the number of particles, the number of iterations, and the learning factor. By setting the target parameter range, it is possible to avoid being trapped in the dilemma of local optimal solutions, which is beneficial to improving the accuracy and comprehensiveness of the prediction of the PSO-SVR effluent water quality prediction model.

[0047] Adopt the K-fold cross-validation method for model verification, and use indicators such as the coefficient of determination (R²), Nash-Sutcliffe efficiency coefficient (NSE), root mean square error (RMSE), and mean absolute percentage error (MAPE) to comprehensively evaluate the model performance.

[0048] In the second aspect, the present invention also provides a treatment device, which adopts the overflow sewage coagulation-oxidation treatment method based on support vector regression and particle swarm optimization algorithm as described in the above solution. The treatment device includes: a reaction tank, a sedimentation tank, and a water quality analyzer. One side of the reaction tank is connected to the sedimentation tank through a pipeline, and the sedimentation tank is connected to the water quality analyzer. The water quality analyzer is used to analyze the concentrations of SS, COD, TN, NH3-N, and TP in the supernatant from the sedimentation tank.

[0049] Compared with the prior art, the beneficial effects of the present invention are:

[0050] 1. By comprehensively considering the co-removal of multiple pollutants and using the support vector regression (SVR) and particle swarm optimization algorithm (PSO) to optimize the coupling process parameters of sodium hypochlorite (NaClO) oxidation and polyaluminum chloride (PACl) coagulation, the present invention can effectively improve the treatment effect of urban overflow sewage, ensuring water environment safety and sustainable development.

[0051] 2. By optimizing the coagulation and oxidation processes, the present invention can effectively and synergistically remove various pollutants in the overflow sewage, including SS, COD, TOC, TN, NH3-N, and TP, significantly improving the water quality. At present, there is little research on the control process for the synergistic removal of pollutants such as carbon, nitrogen, and phosphorus in overflow sewage. The present invention fills this gap and improves the water environment quality.

[0052] 3. By precisely controlling the coagulation-oxidation sequence and the dosing amounts of the coagulant and oxidant, the present invention can save the sewage treatment cost, thereby enhancing the overall capacity of urban sewage treatment. It can treat the overflow sewage with different load composite pollution, flexibly adapt to the changes in different pollutant concentrations and water quality characteristics, and has excellent adaptability and practicability, effectively coping with various complex water quality challenges in practical applications.

[0053] 4. Through the prediction model established by using SVR and PSO, the present invention can accurately predict the effluent quality after the treatment of overflow sewage, providing a scientific basis for subsequent process adjustment; it can provide new ideas and methods for future water treatment technologies, promoting the technological progress in the field of sewage treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a graph of the pollutant removal rates of various sewage treatment chemical addition methods of the present invention;

[0055] Figure 2 It is a flowchart of the method framework of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0056] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0057] Please refer to Figure 1-2 , in an embodiment of the present invention, an overflow sewage coagulation-oxidation treatment method based on support vector regression and particle swarm optimization algorithm is used to control the concentration of pollutants in the overflow sewage. The method includes:

[0058] S1: Select the optimal treatment method for the overflow sewage from various sewage treatment chemical addition methods, which can achieve the pretreatment of the overflow sewage. By performing a correlation analysis on various characteristic variables affecting the effluent quality and using a standardization method to eliminate the dimension difference, the model convergence speed can be improved;

[0059] S2: Under the conditions of the optimal treatment method, measure the effluent pollutant concentration under the influence of a single characteristic variable respectively, and judge the treatment effect according to the correlation between the pollutant concentration and the water quality cleanliness;

[0060] S3: Build a prediction model for the water quality index (WQI) of the effluent based on the pollutant concentration.

[0061] Preferably, the pollutants are SS, COD, TN, NH3-N, and TP. Take the overflow sewage from the sewage pipe, and its water quality characteristics are: SS = 396 mg / L, NH3-N = 21.14 mg / L, TN = 22.52 mg / L, COD = 302.5 mg / L, TP = 1.32 mg / L, pH = 8.1, and water temperature = 25 °C.

[0062] Preferably, the method for selecting the optimal treatment method for the overflow sewage from various sewage treatment chemical addition methods includes:

[0063] Take the to-be-treated overflow sewage in a reaction vessel, add sewage treatment chemicals to the reaction vessel in a certain order to make the overflow sewage stratified;

[0064] Calculate the water quality pollution index (WQI) of the overflow sewage to obtain the result data of the overflow sewage;

[0065] Compare all the result data pairwise, select the result data corresponding to the smallest water quality pollution index (WQI) as the optimal result data, and take the treatment method that obtains the optimal result data as the optimal treatment method. By comparing and analyzing the influence of different chemical addition sequences on the treatment results of the overflow sewage, the optimal chemical addition sequence and the treatment method of the chemicals can be quickly judged, which is beneficial to improving the water treatment effect and treatment efficiency of the overflow sewage.

[0066] Preferably, the first water treatment chemical is a coagulant. By using a coagulant as the first water treatment chemical, the tiny particles in the sewage can be aggregated into flocs through chemical or physical effects, which is convenient for subsequent sedimentation and filtration treatment. The second water treatment chemical is an oxidant, and the oxidant can oxidize the COD in the sewage, which is beneficial to removing the organic substances in the sewage through oxidation, improving the odor in the sewage, and reducing the chromaticity of the sewage through the oxidation of the organic pigments in the sewage.

[0067] Preferably, the coagulant is a polyaluminum chloride (PACl) solution with a concentration of 250 mg / L, and the oxidant is a sodium hypochlorite (NaClO) solution with a concentration of 200 mg / L.

[0068] Preferably, the method for calculating the water quality pollution index (WQI) of the overflow sewage to obtain the result data of the overflow sewage includes:

[0069] Using the pollution index calculation formula Calculate the water quality pollution index, where i is the type of pollutant; I i is the water quality pollution index of pollutant i; C i is the concentration of pollutant i; S i is the evaluation standard of pollutant i. The evaluation standard for single indicators refers to the third-level standard in the current national standard for water bodies, "Discharge Standard of Pollutants for Municipal Wastewater Treatment Plants" (GB 18918-2002). The smaller the WQI value, the higher the indication of water quality cleanliness.

[0070] Preferably, the method for adding sewage treatment agents to the reaction vessel in a certain order includes:

[0071] Carry out a synchronous coagulation-oxidation treatment experiment on the overflow sewage: Add the first water treatment agent and the second water treatment agent to the first reaction vessel simultaneously to obtain the first solution; constitute the treatment method of synchronous coagulation-oxidation;

[0072] The method for stratifying the overflow sewage includes:

[0073] After stirring the first solution at a speed of 250 r / min for 30 s, then stir the first solution at a speed of 50 r / min for 4.5 min;

[0074] Let the stirred first solution stand for sedimentation for 5 min to obtain the first stratified solution;

[0075] Preferably, the method for calculating the water quality pollution index (WQI) of the overflow sewage to obtain the result data of the overflow sewage includes:

[0076] Measure the pollutant concentration values in the first stratified solution, as Figure 1 shown, the SS removal rate is 82.33%, the COD removal rate is 0%, the TN removal rate is 79.77%, the NH3-N removal rate is 77.53%, and the TP removal rate is 85%;

[0077] Based on the pollutant concentration, use the pollution index calculation formula to calculate the water quality pollution index (WQI) of the first stratified solution, and obtain the first pollution index according to the calculation result.

[0078] Preferably, the method for adding sewage treatment agents to the reaction vessel in a certain order to stratify the overflow sewage further includes: [[ID=4,]]

[0079] Conduct an experiment on pre - coagulation - oxidation treatment of overflow sewage: Add the first water treatment agent to the second reaction vessel to obtain the second solution, stir the second solution at a speed of 250 r / min for 1 min, and after the stirring ends, let the second solution stand for 5 min;

[0080] Add the second water treatment agent to the second solution after standing, stir the second solution with the added second water treatment agent at a speed of 100 r / min for 5 min, and after the stirring ends, let the second solution stand and precipitate for 5 min to obtain the second stratified solution;

[0081] The method for calculating the water quality pollution index (WQI) of overflow sewage to obtain the result data of overflow sewage further includes:

[0082] Measure the pollutant concentration in the second stratified solution, as Figure 1 shown, the SS removal rate is 37.90%, the TN removal rate is 11.05%, the NH3 - N removal rate is 20.70%, the COD removal rate is 21.2%, and the TP removal rate is 86%;

[0083] Based on the pollutant concentration, use the pollution index calculation formula to calculate the water quality pollution index (WQI) of the second stratified solution, and obtain the second pollution index according to the calculation result.

[0084] Preferably, the method for adding sewage treatment agents to the reaction vessel in a certain order to stratify the overflow sewage further includes:

[0085] Conduct an experiment on pre - oxidation - coagulation treatment of overflow sewage. Add the second water treatment agent to the third reaction vessel to obtain the third solution, stir the third solution at a speed of 50 r / min for 5 min, and add the first water treatment agent to the stirred third solution; constituting a pre - oxidation - coagulation treatment method;

[0086] After stirring at a speed of 250 r / min for 1 min, let it stand and precipitate for 5 min to obtain the third stratified solution;

[0087] The method for calculating the water quality pollution index (WQI) of overflow sewage to obtain the result data of overflow sewage further includes:

[0088] Measure the pollutant concentration in the third stratified solution, as Figure 1 shown, the SS removal rate is 99.37%, the TN removal rate is 85.74%, the NH3 - N removal rate is 88.28%, the COD removal rate is 0%, and the TP removal rate is 84%; Based on the pollutant concentration, use the pollution index calculation formula to calculate the water quality pollution index (WQI) of the third stratified solution, and obtain the third pollution index according to the calculation result.

[0089] Calculate the effluent WQI after treating overflow sewage with different coagulation-oxidation sequences, and compare the WQI values after different processes, as shown in Table 1 below:

[0090] Table 1 Effluent WQI values after treating overflow sewage with different coagulation-oxidation sequences.

[0091]

[0092] It is obtained that the WQI value is the lowest when pre-oxidation-coagulation is used to treat overflow sewage. Therefore, pre-oxidation-coagulation is the optimal process for treating overflow sewage.

[0093] Preferably, a single characteristic variable includes any one of the factors such as coagulant concentration value, oxidant concentration value, coagulation time, oxidation time, pH value, water temperature value, and influent water quality data value. By introducing characteristic variables including pH value, water temperature value, and influent water quality, it is beneficial to improve the prediction accuracy of the model by increasing the breadth of the characteristic variables;

[0094] Under the conditions of the optimal treatment method, the methods for measuring the effluent pollutant concentration under the influence of a single characteristic variable include:

[0095] According to the optimal treatment method, each time the value of one factor in the characteristic variables is changed, the pollutant concentration in the effluent is measured to obtain the pollutant concentration value;

[0096] Explore the treatment effects of overflow sewage under different conditions of coagulant concentration, oxidant concentration, coagulation time, oxidation time, pH value, water temperature, and influent water quality respectively;

[0097] Under the conditions of the optimal treatment method, the concentration of the coagulant is controlled at 50 mg / L, 100 mg / L, 150 mg / L, 200 mg / L, 250 mg / L, 300 mg / L, 350 mg / L, and 400 mg / L;

[0098] The concentration of the oxidant is controlled at 50 mg / L, 100 mg / L, 150 mg / L, 200 mg / L, 250 mg / L, 300 mg / L, 350 mg / L, and 400 mg / L;

[0099] The coagulation time is controlled at 30 s, 1 min, 2 min, 3 min, 4 min, and 5 min;

[0100] The oxidation time is controlled at 30 s, 1 min, 3 min, 5 min, 10 min, 15 min;

[0101] The pH value is controlled at 5, 6, 7, 8, 9;

[0102] The water temperature was controlled at 10, 15, 20, 25, 30 °C;

[0103] SS was controlled at 100 mg / L, 200 mg / L, 300 mg / L, 400 mg / L, 500 mg / L and 600 mg / L;

[0104] The concentration of NH3-N was controlled at 10mg / L, 15mg / L, 20mg / L, 25mg / L and 30mg / L;

[0105] The concentration of COD was controlled at 100 mg / L, 200 mg / L, 300 mg / L, 400 mg / L and 500 mg / L;

[0106] The concentration of TP was controlled at 0.5mg / L, 1 mg / L, 1.5 mg / L, 2 mg / L and 2.5 mg / L. Experiments were carried out respectively to analyze the removal effect of conventional pollutants in the overflow sewage, and the effluent WQI index was calculated.

[0107] The method for constructing a prediction model of the effluent water quality index (WQI) based on pollutant concentrations includes:

[0108] The support vector regression (SVR) model was used to predict the effluent water quality indexes, and the effluent water quality indexes were pollutant concentrations. The particle swarm optimization algorithm (PSO) was used to optimize the hyperparameters of the support vector regressor (SVR) model. The hyperparameters included the penalty factor C and the kernel function parameter gamma, and a PSO-SVR effluent water quality prediction model was constructed.

[0109] Specifically, a prediction model of the effluent WQI was established, and the R², NSE, RMSE, and MAPE indexes were used to comprehensively evaluate the model performance. The results are shown in Table 2 below:

[0110] Table 2 Performance evaluation results of the support vector regressor (SVR) model optimized by the particle swarm optimization algorithm (PSO).

[0111]

[0112] R² = 0.881~0.938, NSE = 0.879 ~0.937, RMSE = 2.713~3.752 ng / L, MAPE = 32.85%~36.59%, indicating that the evaluation model has high prediction performance.

[0113] The results show that under the water quality conditions in step (1), when the WQI is controlled at 1, the optimal dosing concentrations of the coagulant and oxidant in the pre-oxidation - coagulation process are 250 mg / L and 232 mg / L respectively, and the optimal times for coagulation and oxidation are 1 min and 5 min respectively.

[0114] Preferably, the method for optimizing the hyperparameters of the support vector regression (SVR) model using the particle swarm optimization (PSO) algorithm includes:

[0115] Set the target parameter range. The parameters include the number of particles, the number of iterations, and the learning factor. By setting the target parameter range, it is possible to avoid being trapped in the dilemma of local optimal solutions, which is beneficial to improving the accuracy and comprehensiveness of the prediction of the PSO-SVR effluent quality prediction model.

[0116] Adopt the K-fold cross-validation method for model validation, and use indicators such as the coefficient of determination (R²), Nash-Sutcliffe efficiency coefficient (NSE), root mean square error (RMSE), and mean absolute percentage error (MAPE) to comprehensively evaluate the model performance. The present invention can provide an efficient coagulation-oxidation treatment scheme for overflow sewage with different water quality characteristics, providing new technical support for the sewage treatment field.

[0117] In a second aspect, the present invention also provides a treatment device that adopts the overflow sewage coagulation-oxidation treatment method based on support vector regression and particle swarm optimization algorithm as described above. The treatment device includes: a reaction tank, a sedimentation tank, and a water quality analyzer. One side of the reaction tank is connected to the sedimentation tank through a pipeline, and the sedimentation tank is connected to the water quality analyzer. The water quality analyzer is used to analyze the concentrations of SS, COD, TN, NH3-N, and TP in the supernatant from the sedimentation tank.

[0118] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, makes equivalent substitutions or changes, and should be covered within the protection scope of the present invention.

Claims

1. A coagulation-oxidation treatment method for overflow sewage based on support vector regression and particle swarm optimization algorithm, characterized in that: The method is used to control the concentration of pollutants in overflow sewage, and the method comprises: The optimal treatment method for overflow sewage is selected from a plurality of methods for adding sewage treatment agents, comprising placing the overflow sewage to be treated in a reaction vessel, adding sewage treatment agents to the reaction vessel in a certain order, and stratifying the overflow sewage, wherein the method for stratifying the overflow sewage comprises: stirring a first solution at a speed of 250 r / min for 30 seconds, and then stirring the first solution at a speed of 50 r / min for 4 to 5 minutes; allowing the stirred first solution to settle for 5 minutes to obtain a first stratified solution; and calculating a water quality index (WQI) of the overflow sewage to obtain result data of the overflow sewage, comprising: measuring a pollutant concentration value in the first stratified solution; Calculate the water quality pollution index (WQI) of the overflow sewage to obtain the result data of the overflow sewage; compare all the result data pairwise, select the result data corresponding to the smallest water quality pollution index (WQI) as the optimal result data, and use the processing method that obtains the optimal result data as the optimal processing method. The method of adding sewage treatment agents to the reaction container in a certain order includes adding a first water treatment agent and a second water treatment agent to the first reaction container at the same time to obtain a first solution; forming a synchronous coagulation-oxidation treatment mode, wherein the first water treatment agent is a coagulant and the second water treatment agent is an oxidant; Based on the pollutant concentration, the water quality index (WQI) of the first layered solution is calculated using the pollution index calculation formula, and a first pollution index is obtained according to the calculation result; Using the pollution index calculation formula Calculate the water pollution index, where i is the type of pollutant; Ii is the water pollution index of pollutant i; Ci is the concentration of pollutant i; Si is the evaluation standard of pollutant i; Under the conditions of the optimal treatment method, the concentration of effluent pollutants under the influence of a single characteristic variable is measured respectively, and the single characteristic variable includes: any one factor among a coagulant concentration value, an oxidant concentration value, a coagulation time, an oxidation time, a pH value, a water temperature value, and an influent water quality data value. By introducing characteristic variables including the pH value, the water temperature value, and the influent water quality, it is beneficial to increase the breadth of the characteristic variables to improve the prediction accuracy of the model; the method of measuring the concentration of effluent pollutants under the influence of a single characteristic variable under the conditions of the optimal treatment method includes: according to the optimal treatment method, each time changing the value of one factor in the characteristic variable, measuring the concentration of pollutants in the effluent, and obtaining a pollutant concentration value; A prediction model for an effluent water quality index (WQI) is constructed based on pollutant concentration. The method for constructing a prediction model for an effluent water quality index (WQI) based on pollutant concentration includes: using a support vector regression model to predict an effluent water quality index, where the effluent water quality index is the pollutant concentration; and using a particle swarm optimization algorithm to optimize the hyperparameters of the vector regression machine model, where the hyperparameters include a penalty factor C and a kernel function parameter gamma, to construct a PSO-SVR effluent water quality prediction model.

2. The overflow sewage coagulation-oxidation treatment method based on support vector regression and particle swarm optimization algorithm according to claim 1 is characterized in that: The pollutants are SS, COD, TN, NH3-N and TP.

3. The overflow sewage coagulation-oxidation treatment method based on support vector regression and particle swarm optimization algorithm according to claim 2 is characterized in that: The coagulant is a polyaluminium chloride solution with a concentration of 250 mg / L, and the oxidant is a sodium hypochlorite solution with a concentration of 200 mg / L.

4. The overflow sewage coagulation-oxidation treatment method based on support vector regression and particle swarm optimization algorithm according to claim 2 is characterized in that: The method of adding sewage treatment agents to the reaction container in a certain order to stratify the overflow sewage also includes: Add the first water treatment agent to the second reaction container to obtain a second solution, stir the second solution at a speed of 250 r / min for 1 minute, and let the second solution stand for 5 minutes after the stirring is completed; adding a second water treatment agent to the second solution after standing, stirring the second solution with the second water treatment agent at a speed of 100 r / min for 5 minutes, and after the stirring is completed, allowing the second solution to settle for 5 minutes to obtain a second layered solution; The method for calculating the water quality index WQI of overflow sewage to obtain result data of overflow sewage also includes: measuring the concentration of pollutants in the second layered solution; Based on the pollutant concentration, the water quality pollution index WQI of the second layered solution is calculated using the pollution index calculation formula, and a second pollution index is obtained according to the calculation result.

5. The overflow sewage coagulation-oxidation treatment method based on support vector regression and particle swarm optimization algorithm according to claim 4 is characterized in that: The method of adding sewage treatment agents to the reaction container in a certain order to stratify the overflow sewage also includes: Adding the second water treatment agent to the third reaction container to obtain a third solution, stirring the third solution at a speed of 50 r / min for 5 minutes, and adding the first water treatment agent to the stirred third solution; After stirring at a speed of 250 r / min for 1 min, the mixture was allowed to settle for 5 min to obtain the third layer solution; The method for calculating the water quality index WQI of overflow sewage to obtain result data of overflow sewage also includes: measuring the concentration of pollutants in the third layer solution; Based on the pollutant concentration, the water quality pollution index WQI of the third layered solution is calculated using the pollution index calculation formula, and a third pollution index is obtained according to the calculation result.

6. The overflow sewage coagulation-oxidation treatment method based on support vector regression and particle swarm optimization algorithm according to claim 5 is characterized in that: Methods for optimizing the hyperparameters of the vector regression machine model using the particle swarm optimization algorithm include: Set target parameter range; The K-fold cross-validation method was used for model validation.

Images