Efficient denitrification sewage treatment system and sewage treatment method

Abstract

The invention relates to the technical field of sewage treatment, in particular to an efficient denitrification sewage treatment system and a sewage treatment method. The efficient denitrification sewage treatment system comprises an adjusting tank, an MABR tank, an AOA tank, an MBR tank, a conveying module, a detection module, a doser and an aeration rate regulator. Sewage is conveyed by the conveying module through the adjusting tank, the MBAR tank, the AOA tank and the MBR tank in sequence, water quality data of the sewage is detected by the detection module, and chemicals are added into the adjusting tank and the MBR tank by the chemical feeder, so that impurities in the sewage are separated according to particle sizes, and a biological membrane is prevented from being blocked by the impurities with large particle sizes; the treatment pressure of membrane aeration separation, aerobic and anaerobic separation and membrane separation is reduced. According to the invention, the aeration rate in the MABR tank is optimized through the aeration rate regulator, so that the optimal cooperation of the denitrification rate and the aeration rate is realized, the denitrification efficiency and the denitrification effect are effectively improved, the treatment cost is reduced, and the water quality is further optimized.

Description

Technical Field

[0001] The present invention relates to the technical field of sewage treatment, and in particular to a high-efficiency denitrification sewage treatment system and a high-efficiency denitrification sewage treatment method. Background Art

[0002] Sewage denitrification is of great significance, including protecting the water environment, improving water quality, and promoting the health of the ecosystem. For some manufacturers, sewage denitrification treatment also needs to comply with regulatory requirements. Therefore, denitrification treatment of wastewater is particularly important.

[0003] Among the existing technologies, the existing technologies of efficient denitrification sewage treatment systems mainly include the following: gas phase biological denitrification: using nitrifying bacteria under aerobic conditions to convert ammonia nitrogen into nitrate, and then using denitrifying bacteria under anoxic conditions to convert nitrate into nitrogen gas, and finally achieving nitrogen removal. Sequencing batch reactor (SBR): suitable for small-scale sewage treatment and strong flexibility. Membrane bioreactor (MBR): combining membrane filtration with biological treatment, it improves the treatment effect of sewage and can recycle water resources. Rotating disc biological contact oxidation method: nitrogen removal is carried out through the formation and peeling of biofilm, and the operation is simple. Denitrification technology combines different processes, such as A2 / O process: combining anaerobic, anoxic and aerobic processes to achieve efficient denitrification.

[0004] Although these denitrification processes have strong denitrification effects when facing sewage from a single source with light pollution, or have high economic value when the water quality requirements for discharge are not high, it is difficult for a single sewage treatment method to achieve the ideal denitrification effect when the sources are complex or the water quality requirements for water recycling are high. For example, when using a membrane bioreactor to treat sewage, large particles of impurities can easily cause clogging of membrane pores, reduce membrane flux, and thus reduce sewage treatment efficiency. Or, when using the AO process, although the combination of aerobic and anaerobic treatment processes has a high efficiency in denitrification and dephosphorization, it is difficult to discharge other impurities in the sewage. As the sewage treatment proceeds, the microbial environment of the AO process is difficult to maintain, and the continuous AO treatment process has a high operating cost. Summary of the invention

[0005] Based on this, it is necessary to propose an efficient denitrification sewage treatment system and sewage treatment method to address the problems of low denitrification efficiency, unsatisfactory denitrification effect and high cost of denitrification process in the existing sewage denitrification treatment process.

[0006] The present invention is achieved through the following technical scheme: A high-efficiency denitrification sewage treatment system includes: a regulating tank, a MABR tank, an AOA tank, an MBR tank, a conveying module, a detection module, a doser and an aeration rate regulator.

[0007] The equalization tank uses sedimentation method to perform preliminary separation of sewage and remove large particle impurities in the sewage.

[0008] The MABR pool uses microbubble aeration to provide oxygen, and performs nitrification on the sewage, intercepting microorganisms and suspended matter in the sewage.

[0009] AOA pools include anaerobic pools, aerobic pools, and anoxic pools. Anaerobic pools use anaerobic microorganisms to ferment organic matter and reduce sulfate and other oxidants to form carbon dioxide and ammonia. Aerobic pools use aerobic microorganisms to oxidize ammonia nitrogen compounds in sewage into nitrites and nitrates, and use denitrifying microorganisms to reduce nitrates to nitrogen gas. Anoxic pools use anaerobic microorganisms to decompose organic matter in sewage, release ammonia nitrogen compounds, and convert ammonia nitrogen compounds into gaseous nitrogen through the denitrification process. At the same time, phosphate bacteria are used to release phosphorus in an anaerobic environment.

[0010] The MBR pool purifies sewage water through membrane separation technology, and uses a backwash pump to clean the ultrafiltration membrane in the MBR pool to maintain the membrane flux of the ultrafiltration membrane.

[0011] The conveying module is used to introduce sewage into the equalization tank, and convey the initially separated clean water along the equalization tank, MABR tank, AOA tank and MBR tank in sequence.

[0012] The detection module is used to detect the water quality data of sewage.

[0013] The doser is used to add corresponding types and weights of chemicals to the equalization tank and MBR tank according to water quality data.

[0014] The aeration rate regulator is used to set the optimal aeration rate according to water quality data, and then adjust the aeration rate of the MABR pool according to the optimal aeration rate.

[0015] The present invention sequentially passes sewage through a regulating tank, a MBAR tank, an AOA tank, and an MBR tank, so that the sewage sequentially undergoes chemical addition sedimentation separation, membrane aeration separation, anaerobic separation, and membrane separation, so that impurities of different particle sizes in the sewage are separated according to their particle sizes, thereby avoiding blockage of the biofilm by large particle size impurities, and reducing the processing pressure of membrane aeration separation, anaerobic separation, and membrane separation. In addition, part of the activated sludge generated during membrane aeration separation and membrane separation can be transported or refluxed to the AOA tank to maintain the microbial concentration in the AOA tank, ensure sewage treatment efficiency, and reduce operating costs.

[0016] Furthermore, the detection module includes a sewage initial data detection unit, a separation flow rate detection unit, an aeration rate detection unit, and a water quality detection unit. The sewage initial data detection unit is used to detect the initial water quality data of the sewage. The separation flow rate detection unit is used to detect the clean water flow rate data, including the rate v1 of clean water output from the regulating tank to the MABR tank, the rate v2 of clean water output from the MABR tank to the AOA tank, the rate v3 of clean water output from the AOA tank to the MBR tank, and the rate v4 of clean water output from the MBR tank to the clean water tank. The aeration rate detection unit is used to monitor the real-time aeration rate r in the MABR tank. The water quality detection unit is used to detect the clean water quality data output by the MABR tank, the AOA tank, and the MBR tank.

[0017] Furthermore, the doser includes a reagent proportioning unit and a reagent mixing and conveying unit. The reagent proportioning unit is used to query the reagent type in a pre-stored comparison table according to the clean water quality data, and calculate the reagent dosage and the reagent adding rate according to the clean water flow rate data. The reagent mixing and conveying unit is used to mix the reagents according to the reagent type and dosage to obtain a mixed liquid, and convey the mixed liquid according to the reagent adding rate.

[0018] Furthermore, the mixed liquid medicine proportioning unit constructs an LSTM model optimized by the sparrow algorithm and inputs the historical water quality data into the LSTM model to obtain water quality prediction data. The specific process includes: Collect historical water quality data and standardize them.

[0019] Build an LSTM model for predicting water quality changes.

[0020] The sparrow algorithm is used to optimize the hyperparameters of the LSTM model.

[0021] The historical water quality data are input into the optimized LSTM model to obtain the water quality prediction data.

[0022] Furthermore, the specific steps of using the sparrow algorithm to optimize the hyperparameters of the LSTM model include: The hyperparameters of the LSTM model are encoded as the position vector of each individual sparrow.

[0023] Apply the search mechanism of the sparrow algorithm to calculate the new position of each sparrow individual.

[0024] Select the sparrow individual with the highest fitness as the leader and the other sparrow individuals as followers.

[0025] Determine whether the termination condition is met. If so, output the hyperparameters corresponding to the leader with the highest fitness. Otherwise, update the positions of the leader and followers, and reselect the leader based on the fitness of each sparrow individual after the update.

[0026] Furthermore, the leader's position update formula is expressed as: ;

[0027] In the formula, is the position of the i-th sparrow at the t-th iteration, X b is the position of the current optimal solution, α is the step size control parameter, and R is a random number.

[0028] The follower's position update formula is expressed as: ; In the formula, is the position of the current worst solution, and β is a random number.

[0029] The calculation formula of fitness is expressed as: ; In the formula, y i is the actual value of the water quality data, and n is the number of samples.

[0030] Furthermore, the aeration rate regulator includes an optimal aeration rate simulation unit, an optimal aeration rate calculation unit, and an aeration control unit. The optimal aeration rate simulation unit is used to construct an optimal aeration rate analysis model based on the LSTM network improved by the jumping spider algorithm. The optimal aeration rate calculation unit is used to input water quality data and clean water flow rate data into the optimal aeration rate analysis model and output the optimal aeration rate. The aeration control unit is used to aerate the MABR pool according to the optimal aeration rate.

[0031] Furthermore, the construction process of the optimal aeration rate analysis model includes: A denitrification simulation database is constructed, and the denitrification simulation database is cleaned and normalized to obtain preprocessed data.

[0032] Construct a basic aeration rate analysis model based on LSTM network.

[0033] The hyperparameters of the basic aeration rate analysis model were optimized using the jumping spider algorithm to obtain the optimal aeration rate analysis model.

[0034] The optimal aeration rate analysis model is trained using preprocessed data, and the model parameters that ultimately meet the test accuracy are retained.

[0035] Furthermore, the specific process of optimizing the hyperparameters of the optimal aeration rate analysis model using the jumping spider algorithm includes: Initialization: Set the population size and number of iterations of the jumping spider algorithm.

[0036] Individual encoding: The optimal aeration rate analysis model hyperparameter combination is encoded as the jumping spider's position.

[0037] Search process: Update the position of each jumping spider by imitating the spider's jumping and foraging behavior. The position update formula is expressed as: ; In the formula, S i (t) is the position of the ith jumping spider at the tth iteration, ΔS i (t) is the change in position.

[0038] The change in position is expressed as: ; In the formula, S b (t) is the current global optimal position, R i (t) is a random number that follows a certain distribution and is used to simulate the jumping behavior of a spider.

[0039] Evaluation process: Update the position of the candidate solution according to the position and fitness of each jumping spider. Introduce an exploration strategy to randomly generate new candidate solutions in the search space. The fitness of the jumping spider algorithm is expressed as: ; In the formula, ey j is the actual denitrification efficiency of the MABR pool, is the denitrification efficiency predicted by the LSTM model, and m is the number of samples.

[0040] The search and evaluation processes are cycled until the set termination condition is reached, and the hyperparameter combination corresponding to the jumping spider with the highest fitness is output.

[0041] The present invention also provides a method for treating wastewater with high efficiency denitrification, comprising the following steps: S1: Introduce wastewater into the regulating tank. Add reagents to the regulating tank and use sedimentation method for preliminary separation to remove large particle impurities in the sewage. The reagent addition method is as follows: detect wastewater quality data, and predict water quality data based on the LSTM model optimized by the sparrow algorithm. Query the type of reagent in the pre-stored comparison table based on the water quality data and calculate the dosage of the reagent.

[0042] S2: The clean water after preliminary separation is transported to the MABR pool, and microbubble aeration is used to provide oxygen, and the sewage is nitrified to intercept microorganisms and suspended solids in the sewage. The aeration rate setting method is as follows: Build and train the optimal aeration rate analysis model of the jumping spider algorithm to optimize the LSTM network. According to the water quality data and clean water flow rate data, the optimal aeration rate is output in the optimal aeration rate analysis model.

[0043] S3: The clean water separated from the MABR pool is transported to the AOA pool, and denitrification treatment is carried out through the anaerobic pool, aerobic pool and anoxic pool in sequence.

[0044] S4: The clean water separated from the AOA pool is transported to the MBR pool, the water quality is purified by membrane separation technology, and the ultrafiltration membrane in the MBR pool is cleaned by a backwash pump to maintain the membrane flux of the ultrafiltration membrane.

[0045] S5: The clean water that meets the water quality standards is transported to the clean water tank, and the sludge in the regulating tank, MABR tank, AOA tank and MBR tank is discharged to the sludge tank.

[0046] Compared with the prior art, the present invention has the following beneficial effects: The present invention sequentially passes sewage through a regulating tank, a MBAR tank, an AOA tank, and an MBR tank, so that the sewage sequentially undergoes chemical addition sedimentation separation, membrane aeration separation, anaerobic separation, and membrane separation, so that impurities of different particle sizes in the sewage are separated according to their particle sizes, thereby avoiding blockage of the biofilm by large particle size impurities, and reducing the processing pressure of membrane aeration separation, anaerobic separation, and membrane separation. In addition, part of the activated sludge generated during membrane aeration separation and membrane separation can be transported or refluxed to the AOA tank to maintain the microbial concentration in the AOA tank, ensure sewage treatment efficiency, and reduce operating costs.

[0047] The sparrow algorithm of the present invention optimizes the LSTM model. When the hyperparameter range is wide, the sparrow algorithm can effectively avoid falling into the local optimum, thereby finding a better combination, and when multiple performance indicators need to be optimized at the same time, the flexibility and adaptability of the sparrow algorithm can better cope with these complex situations, and the application in large-scale data sets and complex models is more efficient. The sparrow algorithm optimizes the LSTM model to match the corresponding dosage of the reagent according to the prediction results of the water quality data, realizes the optimization of the sewage denitrification treatment, effectively improves the efficiency and quality of sewage treatment, and reduces the cost of the reagent.

[0048] The present invention uses the LSTM model optimized by the jumping spider algorithm to determine the optimal aeration rate of the MABR pool, so as to optimize the sewage denitrification treatment process, achieve the optimal coordination of the denitrification rate and the aeration rate, effectively improve the denitrification efficiency, and further optimize the water quality. The jumping spider algorithm can effectively optimize the LSTM model, and is suitable for fine-tuning parameters when it is close to the optimal solution. For example, by optimizing within a specific learning rate range, the performance of the LSTM model can be quickly improved. By performing a deep search near the local solution, the jumping spider algorithm can significantly improve the final performance of the model in some cases, and help find the best hyperparameters suitable for a specific data set. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 This is a schematic structural diagram of a high-efficiency denitrification wastewater treatment system according to Example 1 of the present invention; Figure 2 This is a framework diagram of a high-efficiency denitrification wastewater treatment system according to Example 1 of the present invention; Figure 3This is a step diagram of the efficient denitrification wastewater treatment method of Example 2 of the present invention.

[0050] In the figure: 1. Equalization tank; 2. MABR tank; 3. AOA tank; 31. Anaerobic tank; 32. Aerobic tank; 33. Anoxic tank; 4. MBR tank; 5. Clean water tank; 6. Transport module; 7. Aeration rate regulator; 8. Dosing device. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0052] It should be noted that when a component is referred to as being "mounted on" another component, it may be directly on the other component or there may be a central component. When a component is considered to be "set on" another component, it may be directly set on the other component or there may be a central component at the same time. When a component is considered to be "fixed to" another component, it may be directly fixed on the other component or there may be a central component at the same time.

[0053] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which the present invention belongs. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "or / and" used herein includes any and all combinations of one or more of the related listed items.

[0054] Example 1: Please refer to Figure 1 The present embodiment provides a high-efficiency denitrification wastewater treatment system, including a regulating tank 1, a MABR (membrane aerated biofilm reactor) tank 2, an AOA (anaerobic aerobic process) tank 3, an MBR (membrane bioreactor) tank 4, a conveying module 6, a detection module (not marked in the figure), a doser 8 and an aeration rate regulator 7.

[0055] The regulating tank 1 uses a sedimentation method to perform preliminary separation of sewage and remove large particle impurities in the sewage. In this embodiment, the sewage is first introduced into the regulating tank 1, and a sewage lifting pump and a sludge pump are installed in the regulating tank 1 to transport the clean water and sludge after preliminary separation respectively. In addition, a high liquid point and a low liquid point are respectively set in the regulating tank 1, so that the total amount of sewage in the regulating tank 1 is always lower than the high liquid point, so as to avoid the sewage exceeding the regulating tank 1 and causing damage to other equipment, and the precipitated sludge is always lower than the low liquid point to keep the sludge flowing and prevent the sludge from aggregating for a long time. At the same time, when the sewage liquid level is lower than the low liquid point, the sewage continues to be introduced to effectively balance the inlet and outlet water flow.

[0056] In practical applications, a proper amount of reagents can also be added to the regulating tank 1 through the doser 8 to increase the sludge settling rate. For example, flocculants or coagulants such as polyaluminium chloride (PAC) can be added to the regulating tank 1 to remove suspended matter and colloidal substances in sewage. Or denitrification reagents such as sodium hydroxide (for pH adjustment) or specific denitrification agents can be added to the regulating tank 1, depending on the nitrogen and ammonia content. Or dephosphorization agents such as aluminium salts or iron salts can be added to the regulating tank 1 to remove phosphorus from water.

[0057] MABR pool 2 uses microbubble aeration to provide oxygen, and performs nitrification reaction on sewage, intercepting microorganisms and suspended matter in sewage. In this embodiment, the sewage that has undergone preliminary separation first enters MABR pool 2 and contacts with microorganisms on the membrane surface. The pores of the membrane allow oxygen to pass through, providing the oxygen required by microorganisms and promoting the growth of aerobic microorganisms. These microorganisms will oxidize ammonia nitrogen (NH3-N) and convert it into nitrite (NO2⁻) and nitrate (NO3⁻), respectively. At the same time, the organic matter in the sewage provides energy for anaerobic microorganisms, which can convert nitrates into nitrogen gas (N2) to achieve nitrogen removal. As a separation medium, the membrane can effectively separate the treated water from the microorganisms and further improve the water quality.

[0058] The AOA pool 3 includes an anaerobic pool 31, an aerobic pool 32 and an anoxic pool 33; the anaerobic pool 31 mainly performs the following reactions through the action of anaerobic microorganisms: fermentation of organic matter to produce organic acids, hydrogen, etc. Sulfate and other oxidants are reduced to form carbon dioxide and ammonia, etc. Aerobic pool 32 uses aerobic microorganisms to oxidize ammonia nitrogen in sewage into nitrite and nitrate, and uses denitrifying microorganisms to reduce nitrate to nitrogen. The anoxic pool 33 uses anaerobic microorganisms to decompose organic matter in sewage, release ammonia nitrogen, and convert ammonia nitrogen into gaseous nitrogen through the denitrification process. At the same time, in an anaerobic environment, phosphate bacteria will release phosphorus.

[0059] There are still some microorganisms in the clean water after separation in MABR tank 2. This part of clean water is sequentially passed through anaerobic tank 31, aerobic tank 32, and anoxic tank 33 for denitrification treatment, while removing phosphorus therein to reduce the ammonia nitrogen content and total phosphorus content in the clean water.

[0060] In the AOA pool 3, by adjusting the hydraulic retention time, temperature, and the dissolved oxygen content in the aerobic pool 32, the denitrification time is effectively controlled, the denitrification efficiency is improved, and the water quality is further improved.

[0061] MBR pool 4 further purifies the water through membrane separation technology, and uses a backwash pump to clean the membrane to maintain the membrane flux. Specifically, the clean water output from AOA pool 3 is transported to MBR pool 4 and separated by ultrafiltration or microfiltration membrane. The presence of the membrane prevents solid matter (such as activated sludge) from passing through, while allowing clean water to be separated through the membrane to obtain clear effluent. In MBR pool 4, part of the activated sludge will be returned to the aerobic pool 32 to maintain the microbial concentration in the AOA system, ensure treatment efficiency, and reduce operating costs.

[0062] The conveying module 6 is used to introduce sewage into the regulating tank 1, and convey the separated clean water along the regulating tank 1, MABR tank 2, AOA tank 3, and MBR tank 4 in sequence, and discharge the separated sludge into the sludge tank. The conveying module 6 is also used to convey clean water that meets the water quality requirements to the clean water tank 5. In practical applications, due to the different sources of sewage, there are large differences in sewage water quality, and the water quality of clean water obtained by the same sewage treatment method is also different. Therefore, the clean water can be conveyed out at different stages according to the output water quality requirements. For example, for domestic sewage, after denitrification treatment, it is used for watering or flushing toilets, etc., which is significantly different from the water quality requirements for drinking. The water used for flushing toilets often only needs to be treated by MABR or combined with AOA to meet the water quality requirements, while the quality of drinking water often requires higher water quality standards, which need to be combined with MABR, AOA, and MBR, and even aeration operations can be added in the MBR tank 4 to improve the water quality.

[0063] The detection module includes a sewage initial data detection unit, a separation flow rate detection unit, an aeration rate detection unit, and a water quality detection unit. The sewage initial data detection unit is used to detect the initial water quality data of the raw sewage, including COD (chemical oxygen demand), ammonia nitrogen content, total phosphorus content, pH value, total flow rate, etc. The separation flow rate detection unit is used to detect the rate v1 of clear water output from the regulating tank 1 to the MABR tank 2, the rate v2 of clear water output from the MABR tank 2 to the AOA tank 3, the rate v3 of clear water output from the AOA tank 3 to the MBR tank 4, and the rate v4 of clear water output from the MBR tank 4 to the clear water tank 5. The aeration rate detection unit is used to monitor the real-time aeration rate in the MABR tank 2. The water quality detection unit is used to detect the water quality data of the clear water output by the MABR tank 2, the AOA tank 3, and the MBR tank 4.

[0064] The doser 8 is used to add corresponding types and weights of reagents to the regulating tank 1 and the MBR tank 4 according to the detected clean water quality data. The doser 8 includes a reagent proportioning unit and a reagent mixing and conveying unit. The reagent proportioning unit is used to query the type of reagent in the pre-stored comparison table according to the water quality data, and calculate the amount of reagent and the rate of adding the reagent according to the clean water flow rate and flow velocity.

[0065] In this embodiment, the reagent proportioning unit optimizes the LSTM (Long Short-Term Memory Network) model through the sparrow algorithm to predict water quality changes and determine the corresponding reagent dosage. The specific process is as follows: Historical water quality data is collected by simulating the water quality agent ratio experiment and standardized. Among them, the water quality data includes COD, ammonia nitrogen content, total phosphorus content, pH value, total flow, etc. According to the water quality data, the corresponding agent type is queried in the preset agent comparison table. In this embodiment, the input water quality data is standardized so that its mean is 0 and the variance is 1 to improve the model training effect. At the same time, the water quality data is divided into a training set, a validation set, and a test set. Specifically, it can be divided into 70% training set, 15% validation set, and 15% test set.

[0066] Based on the collected data, a LSTM model is constructed to predict water quality changes. The basic model of LSTM is expressed as: ; In the formula, is the predicted value of water quality data, x is the input value of water quality data, and b is the bias term. In this embodiment, water quality data can be expressed as X={x cod , x nh , x p , x ph , x q}, where x cod 、x nh 、x p 、x ph、x q They respectively represent the COD, ammonia nitrogen content, total phosphorus content, pH value and total flow rate of water quality.

[0067] The sparrow algorithm is used to optimize the hyperparameters of the LSTM model. Specifically, the search mechanism of the sparrow algorithm is applied to calculate the new position of each individual. Excellent individuals are selected and retained according to fitness, and the positions of poor individuals in the group are randomly updated. The fitness of the updated individuals is calculated again, and a new hyperparameter combination is selected. The above update and evaluation process is repeated until the set number of iterations is reached or the early termination condition is met.

[0068] The specific steps are as follows: Initialization: Set the population size, number of iterations, search space range and other parameters of the sparrow algorithm.

[0069] Individual encoding: Encode the hyperparameters of the LSTM model (such as learning rate, batch size, number of hidden layer nodes, etc.) into the position vector of the sparrow.

[0070] Search process: The position of the individual is updated through the search strategy of the sparrow algorithm, thereby optimizing the hyperparameters of the LSTM model. The position update formula is expressed as: ; In the formula, is the position of the i-th sparrow at the t-th iteration, X b is the position of the current optimal solution, α is the step size control parameter, and R is a random number.

[0071] The update formula of local search is expressed as: ; In the formula, is the position of the current worst solution, and β is a random number.

[0072] Evaluation and selection: The fitness of individual sparrows is evaluated based on the prediction performance of the LSTM model (such as mean square error MSE) and the optimal hyperparameters are selected.

[0073] The prediction error of the LSTM model is calculated as the fitness of the sparrow algorithm using the following formula: ; In the formula, y i is the actual value of the water quality data, and n is the number of samples.

[0074] Model training and validation: Use the optimized LSTM model to analyze and predict water quality data. Use the best hyperparameter settings found to train the final LSTM model. The training process is as follows: Use the training set to train the LSTM model, observe the loss of the training set and the validation set to prevent overfitting. If the validation loss is no longer decreasing during the training process, you can stop training early. After each training cycle, use the validation set to check the performance of the model. Adjust the model parameters based on the validation set loss and the corresponding indicators. During the training process, you can save the best model, and then use this model for testing after the model training is completed, and retain the model parameters that finally meet the test accuracy.

[0075] Determine the dosage of the agent: Calculate and adjust the dosage of the agent based on the prediction results. The final dosage can be expressed as: ; Where Mi is the optimal dosage of each agent, and k is an empirical proportional constant, which can be found in the preset agent comparison table or determined based on specific water quality conditions and agent characteristics.

[0076] This embodiment uses the Sparrow algorithm to optimize the LSTM model. When the hyperparameter range is wide, the Sparrow algorithm can effectively avoid falling into the local optimum, thereby finding a better combination. When it is necessary to optimize multiple performance indicators (such as accuracy, loss, etc.) at the same time, the flexibility and adaptability of the Sparrow algorithm can better cope with these complex situations, and the application in large-scale data sets and complex models is more efficient. The Sparrow algorithm optimizes the LSTM model to match the corresponding dosage of the reagent according to the prediction results of the water quality data, realizes the optimization of sewage denitrification treatment, effectively improves the efficiency and quality of sewage treatment, and reduces the cost of reagent use.

[0077] The medicine mixing and conveying unit is used to mix medicines according to the type and amount of medicines, and to control the conveying of the medicine mixture according to the medicine adding rate. The water required for mixing can be added by tap water, or clean water that meets the water quality standard can be directly added from the clean water tank 5.

[0078] The aeration rate regulator 7 is used to set the optimal aeration rate according to the detected clean water quality data, and then adjust the aeration rate in the MABR pool 2. The aeration rate regulator 7 includes an optimal aeration rate simulation unit, an optimal aeration rate calculation unit, and an aeration control unit. The optimal aeration rate simulation unit is used to simulate the aeration speed, find the optimal denitrification efficiency and denitrification rate and economic value under different water quality conditions, build a denitrification simulation database, and use the LSTM network optimized by the jumping spider algorithm to analyze the denitrification simulation database, and obtain the optimal aeration rate analysis model based on the LSTM network improved by the jumping spider algorithm.

[0079] In sewage denitrification treatment, the denitrification efficiency of MABR pool 2 is affected by the aeration rate. In order to obtain the optimal aeration rate, the jumping spider algorithm can be used to optimize the parameters of the LSTM network model. The specific steps are as follows: Data collection: Collect data related to the MABR denitrification process, such as aeration rate, influent ammonia nitrogen concentration, nitrate nitrogen concentration, dissolved oxygen concentration, etc.

[0080] Data preprocessing: Clean and normalize the collected data to meet the needs of the LSTM model. Divide the normalized data into training sets, validation sets, and test sets to ensure the effectiveness and accuracy of the model optimization process.

[0081] Building LSTM model: A time series-based LSTM model was built to predict the denitrification efficiency of MABR.

[0082]

[0083] In the formula, is the predicted denitrification efficiency, u is the input influent data, and the influent data u can be expressed as U={u x ,u Rv ,u v}, where u x 、u Rv 、u v They represent influent water quality, denitrification rate and influent flow rate respectively, and d is the bias term.

[0084] The jumping spider algorithm is used to optimize the hyperparameters of the LSTM model. The specific process is as follows: Initialization: Set the jumping spider algorithm's population size, number of iterations, and other parameters.

[0085] Individual encoding: Encoding the hyperparameters of the LSTM model as the position of the jumping spider.

[0086] Search process: The jumping spider algorithm updates the position and searches for the optimal parameters by imitating the jumping and foraging behaviors of spiders.

[0087]

[0088] In the formula, S i (t) is the position of the ith jumping spider at the tth iteration, ΔS i (t) is the change in position.

[0089] The change in position is calculated using the following formula: ; In the formula, S b (t) is the current global optimal position, R i(t) is a random number that follows a certain distribution and is used to simulate the jumping behavior of a spider.

[0090] The LSTM model prediction error is calculated as the fitness of the jumping spider algorithm using the following formula: ; In the formula, ey j is the actual denitrification efficiency of MABR pool 2, is the denitrification efficiency predicted by the LSTM model, and m is the number of samples.

[0091] In each iteration, the following operations are performed according to the fitness: the solution with better fitness is selected, and the position of the candidate solution is updated according to its position and fitness. An exploration strategy is introduced to randomly generate new solutions in the search space to increase the diversity of the population.

[0092] The search and evaluation process of the jumping spider algorithm is repeated until the maximum number of iterations is reached or the fitness reaches the expected target. After the training is completed, the candidate solution with the best fitness is selected as the optimal hyperparameter configuration.

[0093] Model training and optimization: Use the LSTM model optimized by the jumping spider algorithm to predict the MABR denitrification efficiency. Use the best hyperparameter settings found to train the final LSTM model. The training process is as follows: Use the training set to train the LSTM model, observe the loss of the training set and the validation set to prevent overfitting. If the validation loss is no longer decreasing during the training process, the training can be stopped early. After each training cycle, use the validation set to check the performance of the model. Adjust the model parameters according to the loss of the validation set and the corresponding indicators. During the training process, the best model can be saved, and then used for testing after the model training is completed, and the model parameters that finally meet the test accuracy are retained.

[0094] The optimal aeration rate calculation unit is used to output the optimal aeration rate in the optimized LSTM model according to the water quality data and the clean water flow rate. The optimal denitrification efficiency is predicted based on the LSTM model optimized by the jumping spider algorithm, and the corresponding aeration rate is determined. The formula is expressed as: ; In the formula, argmax r It means finding the aeration rate that maximizes the predicted nitrogen removal efficiency.

[0095] The jumping spider algorithm can effectively optimize the LSTM model and is suitable for fine-tuning parameters when the optimal solution is close. For example, optimization within a specific learning rate range can quickly improve the performance of the LSTM model. By performing deep searches near local solutions, the jumping spider algorithm can significantly improve the final performance of the model in some cases and help find the best hyperparameters for specific data sets. The LSTM model optimized by the jumping spider algorithm determines the optimal aeration rate based on the predicted results of the MABR denitrification efficiency to optimize the sewage denitrification treatment process, achieve the optimal coordination of the denitrification rate and aeration rate, effectively improve the denitrification efficiency, and further optimize the water quality.

[0096] The aeration control unit is used to aerate the MABR tank 2 according to the optimal aeration rate.

[0097] In other embodiments, in order to further improve the denitrification efficiency of sewage treatment, the MBR pool 4 may be subjected to aeration treatment, and the specific aeration rate may also be output by the LSTM model optimized by the jumping spider algorithm.

[0098] Example 2: Please refer to Figure 2 This embodiment provides a method for treating wastewater with high efficiency denitrification, which can be controlled by the high efficiency denitrification wastewater treatment system of embodiment 1, and includes the following steps: S1: Introduce wastewater into regulating tank 1. Add reagents to regulating tank 1 and use sedimentation method for preliminary separation to remove large particle impurities in sewage. The reagent addition method is as follows: detect wastewater quality data, analyze and predict water quality data based on the optimized LSTM model. Query the type of reagent in the pre-stored comparison table based on the water quality data and calculate the dosage.

[0099] S2: The clean water after preliminary separation is transported to MABR pool 2, and microbubble aeration is used to provide oxygen, and the sewage is nitrified to intercept microorganisms and suspended solids in the sewage. The aeration rate setting method is as follows: Build and train the LSTM model optimized by the jumping spider algorithm. According to the water quality data and the clean water flow rate, the optimal aeration rate is output in the LSTM model optimized by the jumping spider algorithm.

[0100] S3: The clean water separated from the MABR tank 2 is transported to the AOA tank 3, and denitrification treatment is carried out through the anaerobic tank 31, the aerobic tank 32, and the anoxic tank 33 in sequence.

[0101] S4: MBR pool 4 further purifies the water through membrane separation technology, and uses a backwash pump to clean the membrane to maintain membrane flux.

[0102] S5: The clean water that meets the water quality standards is transported to the clean water tank 5, and the sludge in the regulating tank 1, MABR tank 2, AOA tank 3, and MBR tank 4 is discharged to the sludge tank.

[0103] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0104] The above-mentioned embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A highly efficient denitrification wastewater treatment system, characterized in that: include: The regulating tank uses sedimentation method to perform preliminary separation of sewage and remove large particles of impurities in the sewage; The MABR tank uses microbubble aeration to provide oxygen and nitrify the sewage, intercepting microorganisms and suspended solids in the sewage; The AOA pool includes an anaerobic pool, an aerobic pool and an anoxic pool; the anaerobic pool uses anaerobic microorganisms to ferment organic matter and reduce sulfate and other oxidants to form carbon dioxide and ammonia; the aerobic pool uses aerobic microorganisms to oxidize ammonia nitrogen compounds in sewage into nitrites and nitrates, and uses denitrifying microorganisms to reduce nitrates to nitrogen gas; the anoxic pool uses anaerobic microorganisms to decompose organic matter in sewage, release ammonia nitrogen compounds, and convert ammonia nitrogen compounds into gaseous nitrogen through the denitrification process, and at the same time, uses phosphate bacteria to release phosphorus in an anaerobic environment; The MBR pool purifies the sewage water quality by membrane separation technology, and uses a backwash pump to clean the ultrafiltration membrane in the MBR pool to maintain the membrane flux of the ultrafiltration membrane; A conveying module, used for introducing sewage into the regulating tank, and conveying the initially separated clean water along the regulating tank, the MABR tank, the AOA tank, and the MBR tank in sequence; Detection module, used to detect water quality data of sewage; A doser, used for adding corresponding types and weights of agents to the regulating tank and the MBR tank respectively according to the water quality data; An aeration rate regulator is used to set an optimal aeration rate according to the water quality data, and then adjust the aeration rate of the MABR pool according to the optimal aeration rate.

2. A highly efficient denitrification wastewater treatment system according to claim 1, characterized in that: The detection module includes a sewage initial data detection unit, a separation flow rate detection unit, an aeration rate detection unit, and a water quality detection unit; the sewage initial data detection unit is used to detect the initial water quality data of the sewage; the separation flow rate detection unit is used to detect the clean water flow rate data, including the rate v1 of clean water output from the regulating tank to the MABR tank, the rate v2 of clean water output from the MABR tank to the AOA tank, the rate v3 of clean water output from the AOA tank to the MBR tank, and the rate v4 of clean water output from the MBR tank to the clean water tank; the aeration rate detection unit is used to monitor the real-time aeration rate r in the MABR tank; the water quality detection unit is used to detect the clean water quality data output by the MABR tank, the AOA tank, and the MBR tank.

3. A highly efficient denitrification wastewater treatment system according to claim 1, characterized in that: The doser includes a drug proportioning unit and a drug mixing and conveying unit. The drug proportioning unit is used to query the drug type in a pre-stored comparison table according to the clean water quality data, and calculate the drug amount and drug addition rate according to the clean water flow rate data; The medicine mixing and conveying unit is used to mix the medicines according to the medicine types and the medicine amounts to obtain a mixed medicine liquid, and convey the mixed medicine liquid according to the medicine adding rate.

4. A highly efficient denitrification wastewater treatment system according to claim 3, characterized in that: The mixed liquid medicine proportioning unit constructs an LSTM model optimized by the sparrow algorithm and inputs historical water quality data into the LSTM model to obtain water quality prediction data. The specific process includes: Collect historical water quality data and standardize it; Build an LSTM model for predicting water quality changes; The sparrow algorithm is used to optimize the hyperparameters of the LSTM model; The historical water quality data is input into the optimized LSTM model to obtain water quality prediction data.

5. A highly efficient denitrification wastewater treatment system according to claim 4, characterized in that: The specific steps of using the sparrow algorithm to optimize the hyperparameters of the LSTM model include: Encoding the hyperparameters of the LSTM model as the position vector of each individual sparrow; Apply the search mechanism of the sparrow algorithm to calculate the new position of each individual sparrow; Select the sparrow with the highest fitness as the leader and the other sparrows as followers; Determine whether the termination condition is met. If so, output the hyperparameters corresponding to the leader with the highest fitness. Otherwise, update the positions of the leader and followers, and reselect the leader based on the fitness of each sparrow individual after the update.

6. A highly efficient denitrification wastewater treatment system according to claim 5, characterized in that: The leader’s position update formula is expressed as: ； In the formula, It is The position of the sparrow in the tth iteration, X b is the position of the current optimal solution, α is the step size control parameter, and R is a random number; The follower's position update formula is expressed as: ； In the formula, is the position of the current worst solution, β is a random number; The calculation formula of fitness is expressed as: ； In the formula, is the actual value of the water quality data, and n is the number of samples.

7. The high-efficiency denitrification sewage treatment system according to claim 1 is characterized in that: The aeration rate regulator includes an optimal aeration rate simulation unit, an optimal aeration rate calculation unit, and an aeration control unit; the optimal aeration rate simulation unit is used to construct an optimal aeration rate analysis model based on an improved LSTM network using a jumping spider algorithm; the optimal aeration rate calculation unit is used to input the water quality data and the clean water flow rate data into the optimal aeration rate analysis model and output the optimal aeration rate; the aeration control unit is used to aerate the MABR pool according to the optimal aeration rate.

8. A high-efficiency denitrification sewage treatment system according to claim 7, characterized in that: The construction process of the optimal aeration rate analysis model includes: Constructing a denitrification simulation database, and cleaning and normalizing the denitrification simulation database to obtain preprocessed data; Construct a basic aeration rate analysis model based on LSTM network; The jumping spider algorithm is used to optimize the hyperparameters of the basic aeration rate analysis model to obtain the optimal aeration rate analysis model; The optimal aeration rate analysis model is trained using preprocessed data, and model parameters that ultimately meet the test accuracy are retained.

9. The high-efficiency denitrification sewage treatment system according to claim 7, characterized in that: The specific process of optimizing the hyperparameters of the optimal aeration rate analysis model using the jumping spider algorithm includes: Initialization: Set the population size and number of iterations of the jumping spider algorithm; Individual encoding: encoding the hyperparameter combination of the optimal aeration rate analysis model as the position of the jumping spider; Search process: Update the position of each jumping spider by imitating the spider's jumping and foraging behavior; the position update formula is expressed as: ； In the formula, It is The position of a jumping spider at iteration t, is the change in position; The change in position is expressed as: ； In the formula, is the current global optimal position, It is a random number that follows a certain distribution and is used to simulate the jumping behavior of spiders; Evaluation process: Update the position of the candidate solution according to the position and fitness of each jumping spider; introduce an exploration strategy to randomly generate new candidate solutions in the search space; the fitness of the jumping spider algorithm is expressed as: ； In the formula, ey j is the actual denitrification efficiency of the MABR pool, is the denitrification efficiency predicted by the LSTM model, and m is the number of samples; The search process and the evaluation process are looped until a set termination condition is reached, and the hyperparameter combination corresponding to the jumping spider with the highest fitness is output.

10. A method for treating wastewater with high efficiency denitrification, which adopts a system for treating wastewater with high efficiency denitrification as claimed in any one of claims 1 to 9, characterized in that: The following steps are involved: S1: Introduce wastewater into the regulating tank; Add reagents to the regulating tank and use sedimentation method for preliminary separation to remove large particle impurities in the sewage; wherein the reagent adding method is as follows: detect wastewater quality data, and predict the water quality data according to the LSTM model optimized by the sparrow algorithm; query the type of reagent in the pre-stored comparison table according to the water quality data, and calculate the amount of the reagent; S2: The clean water after preliminary separation is transported to the MABR pool, and microbubble aeration is used to provide oxygen, and the sewage is subjected to nitrification reaction to intercept microorganisms and suspended solids in the sewage; wherein, the aeration rate setting method is as follows: constructing and training the optimal aeration rate analysis model of the LSTM network optimized by the jumping spider algorithm; outputting the optimal aeration rate in the optimal aeration rate analysis model according to the water quality data and the clean water flow rate data; S3: The clean water separated from the MABR tank is transported to the AOA tank, and denitrified through the anaerobic tank, aerobic tank, and anoxic tank in sequence; S4: transporting the clean water separated from the AOA pool to the MBR pool, purifying the water quality by membrane separation technology, and using a backwash pump to clean the ultrafiltration membrane in the MBR pool to maintain the membrane flux of the ultrafiltration membrane; S5: transporting clean water that meets the water quality standards to the clean water tank, and discharging the sludge in the regulating tank, the MABR tank, the AOA tank, and the MBR tank to the sludge tank.

Images