High-efficiency denitrification sewage treatment system and sewage treatment method

Abstract

The present application relates to sewage treatment technical field, specifically to a kind of efficient denitrification sewage treatment system and sewage treatment method.Efficient denitrification sewage treatment system includes: adjusting pool, MABR pool, AOA pool, MBR pool, conveying module, detection module, dosing device and aeration rate regulator.Conveying module is sequentially transported with sewage through adjusting pool, MBAR pool, AOA pool, MBR pool, and the water quality data of sewage is detected by detection module, and then dosing device is used to add medicine to adjusting pool, MBR pool, so that impurities in sewage are separated according to particle size, avoid large particle size impurities to cause biological membrane to be blocked, reduce membrane aeration separation, good anaerobic separation, membrane separation treatment pressure.The present application optimizes the aeration rate in MABR pool by aeration rate regulator, realizes the optimal cooperation of denitrification rate and aeration rate, effectively improves denitrification efficiency and denitrification effect, reduces processing cost, further optimizes water quality.

Description

TECHNICAL FIELD

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

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

[0003] In the prior art, the existing technology of high-efficiency denitrification sewage treatment system mainly includes 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, finally realizing the removal of nitrogen. Sequencing batch reactor (SBR): suitable for small-scale sewage treatment, high flexibility. Membrane bioreactor (MBR): combining membrane filtration and biological treatment, improving the treatment effect of sewage and recycling water resources. Rotary biological contact oxidation method: nitrogen removal through the formation and shedding of biological membranes, simple operation. Denitrification technology combined with different processes, such as A2 / O process: combining anaerobic, anoxic and aerobic processes to achieve high-efficiency denitrification.

[0004] Although these denitrification processes have strong denitrification effect in the face of single source, light pollution sewage, or have high economic value in the case of low water quality discharge requirements, but for complex source or high water quality recycling requirements, single sewage treatment method is difficult to achieve ideal denitrification effect. For example, when using membrane bioreactor to treat sewage, large particles impurities are easy to cause membrane hole blockage, reduce membrane flux, and further reduce sewage treatment efficiency. Or, when using AO process, although it combines aerobic and anaerobic treatment processes, it has high efficiency for denitrification and dephosphorization, but it is difficult to discharge other impurities in sewage. With the progress of sewage treatment, the microbial environment of AO process is difficult to maintain, and the continuous AO treatment process has high operating cost. SUMMARY

[0005] Therefore, it is necessary to propose a high-efficiency denitrification sewage treatment system and method in view of 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 application is realized by the following technical scheme: a high-efficiency denitrification sewage treatment system comprises an adjusting tank, an MABR tank, an AOA tank, an MBR tank, a conveying module, a detection module, a chemical feeder and an aeration rate regulator.

[0007] The conditioning tank uses a sedimentation method to preliminarily separate the sewage and remove large-particle impurities in the sewage.

[0008] The MABR tank uses micro-bubble aeration to provide oxygen and perform a nitrification reaction on the sewage, and traps microorganisms and suspended solids in the sewage.

[0009] The AOA tank includes an anaerobic tank, an aerobic tank, and an anoxic tank. The anaerobic tank uses anaerobic microorganisms to ferment organic matter and reduce sulfate and other oxidants to form carbon dioxide and ammonia. The aerobic tank uses aerobic microorganisms to oxidize ammonia-nitrogen compounds in the sewage into nitrite and nitrate, and uses denitrifying microorganisms to reduce the nitrate into nitrogen gas. The anoxic tank uses anaerobic microorganisms to decompose organic matter in the sewage to release ammonia-nitrogen compounds, and converts the ammonia-nitrogen compounds into gaseous nitrogen through a denitrification process. In the anaerobic environment, phosphate bacteria are used to release phosphorus.

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

[0011] The conveying module is used to introduce the sewage into the conditioning tank, and sequentially convey the preliminarily separated clean water along the conditioning tank, the MABR tank, the AOA tank, and the MBR tank.

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

[0013] The dosing device is used to add reagents of corresponding types and weights to the conditioning tank and the MBR tank according to the water quality data.

[0014] The 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 tank according to the optimal aeration rate.

[0015] The sewage is sequentially subjected to the conditioning tank, the MABR tank, the AOA tank, and the MBR tank, so that the sewage is sequentially subjected to dosing sedimentation separation, membrane aeration separation, aerobic-anaerobic separation, and membrane separation. Different particle sizes of impurities in the sewage are separated according to the particle size, so that the large-particle impurities do not block the biological membrane, and the treatment pressure of the membrane aeration separation, the aerobic-anaerobic separation, and the membrane separation is reduced. In addition, part of the activated sludge generated in the membrane aeration separation and the membrane separation process can be conveyed or backflowed to the AOA tank to maintain the concentration of microorganisms in the AOA tank, ensure the sewage treatment efficiency, and reduce the operation cost.

[0016] Further, the detection module comprises sewage initial data detection unit, separation flow rate detection unit, aeration rate detection unit, water quality detection unit. The sewage initial data detection unit is used for detecting the initial water quality data of sewage. The separation flow rate detection unit is used for detecting the clear water flow rate data, including the rate v1 of outputting clear water from the adjustment tank to the MABR tank, the rate v2 of outputting clear water from the MABR tank to the AOA tank, the rate v3 of outputting clear water from the AOA tank to the MBR tank, and the rate v4 of outputting clear water from the MBR tank to the clear water tank. The aeration rate detection unit is used for monitoring the real-time aeration rate r in the MABR tank. The water quality detection unit is used for detecting the clear water quality data output by the MABR tank, the AOA tank and the MBR tank.

[0017] Further, the doser comprises a medicament proportioning unit and a medicament mixing and conveying unit. The medicament proportioning unit is used for querying the medicament type in the pre-stored control table according to the clear water quality data, and calculating the medicament amount and the medicament addition rate according to the clear water flow rate data. The medicament mixing and conveying unit is used for mixing the medicaments to obtain mixed medicament liquid according to the medicament type and the medicament amount, and conveying the mixed medicament liquid according to the medicament addition rate.

[0018] Further, the mixed medicament liquid medicament proportioning unit inputs the historical water quality data into the LSTM model optimized by the sparrow algorithm to obtain water quality prediction data, and the specific process comprises:

[0019] Collecting historical water quality data and performing standardization processing.

[0020] Constructing an LSTM model for predicting water quality changes.

[0021] Optimizing the hyperparameters of the LSTM model by using the sparrow algorithm.

[0022] Inputting the historical water quality data into the optimized LSTM model to obtain water quality prediction data.

[0023] Further, the specific steps of optimizing the hyperparameters of the LSTM model by using the sparrow algorithm comprise:

[0024] Encoding the hyperparameters of the LSTM model as the position vector of each sparrow individual.

[0025] Applying the search mechanism of the sparrow algorithm to calculate the new position of each sparrow individual.

[0026] Selecting the sparrow individual with the highest fitness as the leader and the other sparrow individuals as followers.

[0027] Judging whether the termination condition is met, if yes, outputting the hyperparameters corresponding to the leader with the highest fitness, otherwise, updating the positions of the leader and the followers, and reselecting the leader according to the fitness of each sparrow individual after updating.

[0028] Further, the position update formula of the leader is expressed as:

[0029] ;

[0030] In the formula, is the position of the i th sparrow in the t th iteration, X b is the position of the current optimal solution, a is a step control parameter, and R is a random number.

[0031] The position update formula of the follower is expressed as:

[0032] ;

[0033] In the formula, is the position of the current worst solution, and β is a random number.

[0034] The fitness calculation formula is expressed as:

[0035] ;

[0036] In the formula, y i is the actual value of the water quality data, and n is the sample quantity.

[0037] Further, the aeration rate regulator comprises 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 improved LSTM network of the jumping spider algorithm. The optimal aeration rate calculation unit is used to input the water quality data and the clear 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 tank according to the optimal aeration rate.

[0038] Further, the construction process of the optimal aeration rate analysis model comprises:

[0039] A denitrification simulation database is constructed, and the denitrification simulation database is cleaned and normalized to obtain preprocessed data.

[0040] A basic aeration rate analysis model based on the LSTM network is constructed.

[0041] The hyperparameters of the basic aeration rate analysis model optimized by the jumping spider algorithm are adopted to obtain the optimal aeration rate analysis model.

[0042] The optimal aeration rate analysis model is trained using the preprocessed data, and the model parameters that finally meet the test accuracy are retained.

[0043] Further, the specific process of optimizing the hyperparameters of the optimal aeration rate analysis model by the jumping spider algorithm comprises:

[0044] Initialization: Set the population size and the number of iterations of the jumping spider algorithm.

[0045] Individual encoding: Encode the combination of hyperparameters of the optimal aeration rate analysis model as the position of the jumping spider.

[0046] Search process: Update the position of each jumping spider by simulating the jumping and foraging behavior of spiders. The position update formula is expressed as:

[0047] ;

[0048] In the formula, S i (t) is the position of the i th jumping spider at the t th iteration, ΔS i (t) is the change in position.

[0049] In the formula, ΔS

[0050] ;

[0051] In the formula, S b (t) is the current global optimal position, R i (t) is a random number subject to a certain distribution, used to simulate the jumping behavior of spiders.

[0052] Evaluation process: Update the position of the candidate solution according to the position and fitness of each jumping spider. An exploration strategy is introduced to randomly generate new candidate solutions within the search space. The fitness of the jumping spider algorithm is expressed as:

[0053] ;

[0054] In the formula, ey j is the actual denitrification efficiency of the MABR tank, is the denitrification efficiency predicted by the LSTM model, and m is the number of samples.

[0055] Loop the search process and the evaluation process until the set termination condition is reached, and output the hyperparameter combination corresponding to the jumping spider with the highest fitness.

[0056] The present application also provides a high-efficiency denitrification wastewater treatment method, comprising the following steps:

[0057] S1: Introduce wastewater into a conditioning tank. Add reagents to the conditioning tank and use the sedimentation method for preliminary separation to remove large particulate impurities in the wastewater. The reagent addition method is as follows: detect wastewater quality data, predict the water quality data according to the LSTM model optimized by the sparrow algorithm. According to the water quality data, query the reagent type in the pre-stored control table, and calculate the reagent amount.

[0058] S2: deliver the primary separated clean water to the MABR tank, provide oxygen by micro-bubble aeration, and carry out nitrification reaction on the sewage, and intercept microorganisms and suspended solids in the sewage. The aeration rate setting method is as follows: an optimal aeration rate analysis model of the LSTM network optimized by the jumping spider algorithm is constructed and trained. According to the water quality data and the clean water flow rate data, the optimal aeration rate is output in the optimal aeration rate analysis model.

[0059] S3: deliver the clean water separated by the MABR tank to the AOA tank, and sequentially pass through the anaerobic tank, the aerobic tank and the anoxic tank for denitrification treatment.

[0060] S4: deliver the clean water separated by the AOA tank to the MBR tank, purify the water quality by the membrane separation technology, and clean the ultrafiltration membrane in the MBR tank by the backwashing pump, so as to maintain the membrane flux of the ultrafiltration membrane.

[0061] S5: deliver the clean water meeting the water quality standard to the clean water tank, and discharge the sludge in the conditioning tank, the MABR tank, the AOA tank and the MBR tank to the sludge tank.

[0062] Compared with the prior art, the present application has the following beneficial effects:

[0063] The present application makes the sewage pass through the conditioning tank, the MABR tank, the AOA tank and the MBR tank in sequence, so that the impurities of different particle sizes in the sewage are separated according to the particle size, the large particle size impurities are prevented from blocking the biological membrane, and the treatment pressure of the membrane aeration separation, the aerobic and anaerobic separation and the membrane separation is reduced. In addition, part of the activated sludge generated in the membrane aeration separation and the membrane separation process can be delivered or backflow to the AOA tank, so as to maintain the microorganism concentration in the AOA tank, ensure the sewage treatment efficiency, and reduce the operation cost.

[0064] The sparrow algorithm optimizes the LSTM model, and when the hyperparameter range is wide, the sparrow algorithm can effectively avoid falling into local optimum, so as to find a better combination. 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 according to the water quality data prediction result to match the corresponding reagent dosage, realizes the optimization of sewage denitrification treatment, effectively improves the sewage treatment efficiency and quality, and reduces the reagent use cost.

[0065] This invention employs the jumping spider algorithm to optimize an LSTM model, determining the optimal aeration rate for a MABR (Multi-Active Bioreactor) tank to optimize the wastewater denitrification process. This achieves the optimal balance between denitrification and aeration rates, effectively improving denitrification efficiency and further optimizing water quality. The jumping spider algorithm effectively optimizes LSTM models and is suitable for fine-tuning parameters when the solution is close to optimal, such as optimizing within a specific learning rate range. It can rapidly improve LSTM model performance. By performing a deep search near local solutions, the jumping spider algorithm can significantly improve the final performance of the model in certain situations, helping to find the optimal hyperparameters for a specific dataset. Attached Figure Description

[0066] Figure 1 This is a schematic diagram of the high-efficiency denitrification wastewater treatment system according to Embodiment 1 of the present invention;

[0067] Figure 2 This is a framework diagram of the high-efficiency denitrification wastewater treatment system of Embodiment 1 of the present invention;

[0068] Figure 3 This is a flowchart illustrating the steps of the efficient denitrification wastewater treatment method in Embodiment 2 of the present invention.

[0069] In the diagram: 1. Equalization tank; 2. MABR tank; 3. AOA tank; 31. Anaerobic tank; 32. Aerobic tank; 33. Anoxic tank; 4. MBR tank; 5. Clear water tank; 6. Transport module; 7. Aeration rate regulator; 8. Dosing device. Detailed Implementation

[0070] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0071] It should be noted that when a component is said to be "installed on" another component, it can be directly on the other component or it may be in a component that is centered on it. When a component is said to be "set on" another component, it can be directly set on the other component or it may also be in a component that is centered on it. When a component is said to be "fixed to" another component, it can be directly fixed to the other component or it may also be in a component that is centered on it.

[0072] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in the description herein is for describing particular embodiments only and is not intended to be limiting of the application. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and / or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0073] Embodiment 1: Please refer to Figure 1 The embodiment provides a high-efficiency denitrification wastewater treatment system, which comprises an adjusting tank 1, a MABR (membrane aerated biofilm reactor) tank 2, an AOA (anaerobic and aerobic process) tank 3, an MBR (membrane bioreactor) tank 4, a conveying module 6, a detection module (not marked in the figure), a dosing device 8 and an aeration rate regulator 7.

[0074] The adjusting tank 1 adopts the sedimentation method to preliminarily separate the wastewater and remove large-particle impurities in the wastewater. In the embodiment, the wastewater is first introduced into the adjusting tank 1, and a wastewater lifting pump and a sludge pump are installed in the adjusting tank 1 and used to convey the clear water and sludge after preliminary separation, respectively. In addition, a high liquid level point and a low liquid level point are arranged in the adjusting tank 1, so that the total amount of wastewater in the adjusting tank 1 is always lower than the high liquid level point, damage to other equipment caused by the wastewater overflowing the adjusting tank 1 is avoided, the sludge after sedimentation is always lower than the low liquid level point, so that the sludge can flow and the sludge is prevented from gathering for a long time, and meanwhile, when the liquid level of the wastewater is lower than the low liquid level point, the wastewater is continuously introduced, so that the inflow and outflow of the water are effectively balanced.

[0075] In actual application, an appropriate amount of medicament can also be added to the adjusting tank 1 by the dosing device 8 to improve the sludge settling rate. For example, a flocculating agent or a coagulating agent such as polyaluminum chloride (PAC) can be added to the adjusting tank 1 to remove suspended solids and colloidal substances in the wastewater. Or a denitrification medicament such as sodium hydroxide (used for adjusting pH) or a specific denitrification agent can be added to the adjusting tank 1 according to the nitrogen ammonia content decision. Or a phosphorus removal agent such as an aluminum salt or an iron salt can be added to the adjusting tank 1 to remove phosphorus in the water.

[0076] The MABR tank 2 adopts micro-bubble aeration to provide oxygen and perform a nitrification reaction on the wastewater, and intercept microorganisms and suspended solids in the wastewater. In the embodiment, the wastewater after preliminary separation is first introduced into the MABR tank 2 and contacts the microorganisms on the membrane surface. The pores of the membrane allow oxygen to pass through, providing the oxygen required by the microorganisms and promoting the growth of aerobic microorganisms. These microorganisms will respectively oxidize ammonia nitrogen (NH3-N) and convert it into nitrite (NO2⁻) and nitrate (NO3⁻). At the same time, the organic matter in the wastewater provides energy for anaerobic microorganisms, which can convert nitrate into nitrogen (N2), achieving nitrogen removal. The membrane acts as a separation medium and can effectively separate the treated water from the microorganisms, further improving the water quality.

[0077] The AOA tank 3 includes an anaerobic tank 31, an aerobic tank 32, and an anoxic tank 33. The anaerobic tank 31 mainly performs the following reactions by the action of anaerobic microorganisms: fermentation of organic matter to generate organic acid, hydrogen, etc. Sulfate and other oxidants are reduced to form carbon dioxide and ammonia, etc. The aerobic tank 32 uses aerobic microorganisms to oxidize ammonia nitrogen in the wastewater to nitrite and nitrate, and uses denitrifying microorganisms to reduce nitrate to nitrogen. The anoxic tank 33 uses anaerobic microorganisms to decompose organic matter in the wastewater to release ammonia nitrogen, and converts the ammonia nitrogen to gaseous nitrogen through the denitrification process. Meanwhile, in the anaerobic environment, phosphate bacteria release phosphorus.

[0078] After being separated by the MABR tank 2, there are still some microorganisms in the clear water. This part of the clear water is sequentially subjected to denitrification treatment in the anaerobic tank 31, the aerobic tank 32, and the anoxic tank 33, and the phosphorus therein is removed, thereby reducing the ammonia nitrogen content and total phosphorus content in the clear water.

[0079] In the AOA tank 3, by adjusting the hydraulic retention time, temperature, and the dissolved oxygen content in the aerobic tank 32, the denitrification time is effectively controlled, the denitrification efficiency is improved, and the water quality is further improved.

[0080] The MBR tank 4 further purifies the water quality by membrane separation technology, and uses a backwashing pump to clean the membrane to maintain the membrane flux. Specifically, the clear water output from the AOA tank 3 is transported into the MBR tank 4 and separated by ultrafiltration or microfiltration membranes. The presence of the membrane prevents solid substances (such as activated sludge) from passing through, while allowing clear water to pass through the membrane for separation to obtain clear effluent. In the MBR tank 4, part of the activated sludge is backflowed to the aerobic tank 32 to maintain the concentration of microorganisms in the AOA system, ensure the treatment efficiency, and reduce the operating cost.

[0081] The conveying module 6 is used to introduce the wastewater into the conditioning tank 1, and sequentially convey the separated clear water along the conditioning tank 1, the MABR tank 2, the AOA tank 3, and the MBR tank 4, and discharge the separated sludge to the sludge tank. The conveying module 6 is also used to convey the clear water meeting the water quality requirements to the clear water tank 5. In actual application, due to the difference in the source of the wastewater, the water quality of the wastewater is quite different, and the water quality of the clear water obtained by using the same wastewater treatment method is also different. Therefore, the clear water can be conveyed out at different stages according to the water quality requirements. For example, for domestic wastewater, after denitrification treatment, it is used for watering or flushing, etc., which is obviously different from the water quality requirement for drinking. The water used for flushing often only needs to be treated by MABR or combined with AOA to meet the water quality requirement, while the water quality for drinking often needs to meet higher water quality standards, and needs to be treated by MABR, AOA, and MBR in combination, or even aerated in the MBR tank 4 to improve the water quality.

[0082] The detection module includes an initial wastewater data detection unit, a separation flow rate detection unit, an aeration rate detection unit, and a water quality detection unit. The initial wastewater data detection unit detects the initial water quality data of the raw wastewater, including COD (Chemical Oxygen Demand), ammonia nitrogen content, total phosphorus content, pH value, and total flow rate. The separation flow rate detection unit detects the rates v1 (from equalization tank 1 to MABR tank 2), v2 (from MABR tank 2 to AOA tank 3), v3 (from AOA tank 3 to MBR tank 4), and v4 (from MBR tank 4 to clear water tank 5). The aeration rate detection unit monitors the real-time aeration rate in MABR tank 2. The water quality detection unit detects the water quality data of the clear water output from MABR tank 2, AOA tank 3, and MBR tank 4.

[0083] The dosing device 8 is used to add the appropriate type and weight of chemicals to the equalization tank 1 and the MBR tank 4 according to the detected clean water quality data. The dosing device 8 includes a chemical proportioning unit and a chemical mixing and delivery unit. The chemical proportioning unit is used to look up the type of chemical in a pre-stored reference table according to the water quality data, and to calculate the dosage and chemical addition rate according to the clean water flow rate and velocity.

[0084] In this embodiment, the reagent dosing unit uses the Sparrow Algorithm to optimize the LSTM (Long Short-Term Memory) network model to predict water quality changes and determine the corresponding reagent dosage. The specific process is as follows:

[0085] Historical water quality data was collected through simulated water quality reagent mixing experiments and then standardized. This data included COD, ammonia nitrogen content, total phosphorus content, pH value, and total flow rate. The corresponding reagent type was then looked up in a pre-defined reagent reference table based on the water quality data. In this embodiment, the input water quality data was standardized to have a mean of 0 and a variance of 1 to improve model training effectiveness. The water quality data was also divided into training, validation, and test sets, specifically 70% for training, 15% for validation, and 15% for test.

[0086] Based on the collected data, an LSTM model is constructed to predict water quality changes. The basic LSTM model is expressed as follows:

[0087] ;

[0088] In the formula, Here, is the predicted value of the water quality data, x is the input value of the water quality data, and b is the bias term. In this embodiment, the water quality data can be represented as X = {x...} cod x nh x p x ph x q}, where x cod xnh , x p , x ph , x q COD, ammonia nitrogen content, total phosphorus content, pH value, and total flow, respectively.

[0089] 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. The excellent individuals are selected based on the fitness, and the positions of the poor individuals in the population are randomly updated. The fitness of the updated individuals is calculated again, and a new combination of hyperparameters is selected. The above updating and evaluation process is repeated until the set number of iterations is reached or the early termination condition is met.

[0090] The specific steps are as follows:

[0091] Initialization: Set the population size, number of iterations, search space range, and other parameters of the sparrow algorithm.

[0092] 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.

[0093] Search process: Update the position of the individual through the search strategy of the sparrow algorithm to optimize the hyperparameters of the LSTM model. The position update formula is expressed as:

[0094] ;

[0095] where, is the position of the ith sparrow at the tth iteration, X b is the position of the current optimal solution, and α is the step control parameter and R is a random number.

[0096] The update formula of local search is expressed as:

[0097] ;

[0098] where, is the position of the current worst solution, and β is a random number.

[0099] Evaluation and selection: Evaluate the fitness of the sparrow individual based on the prediction performance of the LSTM model (such as mean square error MSE), and select the optimal hyperparameters.

[0100] The following formula is used to calculate the prediction error of the LSTM model as the fitness of the sparrow algorithm:

[0101] ;

[0102] where, i is the actual value of the water quality data, and n is the number of samples.

[0103] Model training and validation: The optimized LSTM model is used to analyze and predict water quality data. The final LSTM model is trained using the best hyperparameter settings found. The training process is as follows: the training set is used to train the LSTM model, and the loss of the training set and validation set is observed to prevent overfitting. If the validation loss is found to no longer decrease during training, training can be stopped early. After each training cycle, the performance of the model is checked using the validation set. The model parameters are adjusted based on 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. The final model parameters that meet the test accuracy are retained.

[0104] Determining the dosage of the reagent: Based on the prediction results, the dosage of the reagent is calculated and adjusted. The final dosage can be expressed as:

[0105] ;

[0106] where Mi is the optimal input amount of each reagent, and k is an empirical proportion constant that can be queried in a pre-set reagent control table or determined based on specific water quality conditions and reagent characteristics.

[0107] In this embodiment, the sparrow algorithm is used to optimize the LSTM model. When the hyperparameter range is wide, the sparrow algorithm can effectively avoid being trapped in local optima, thereby finding a better combination. When multiple performance indicators (such as accuracy and loss) need to be optimized simultaneously, the flexibility and adaptability of the sparrow algorithm can better cope with these complex situations, and its application in large-scale data sets and complex models is more efficient. The sparrow algorithm optimized LSTM model matches the corresponding reagent dosage based on the water quality data prediction results, achieving optimal wastewater denitrification treatment, effectively improving wastewater treatment efficiency and quality, and reducing reagent usage costs.

[0108] The reagent mixing and delivery unit is used to mix reagents according to the type and amount of reagents, and to control the delivery of the reagent mixture according to the reagent addition rate. The water required for mixing can be added by tap water or directly added from the clean water tank 5 to meet the water quality standards.

[0109] The aeration rate regulator 7 is used to set the optimal aeration rate based on the detected clean water quality data, and then adjust the aeration rate in the MABR tank 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, economic value under different water quality conditions, construct a denitrification simulation database, and analyze the denitrification simulation database using the LSTM network optimized by the jumping spider algorithm to obtain the optimal aeration rate analysis model based on the jumping spider algorithm improved LSTM network.

[0110] In the wastewater denitrification process, the denitrification efficiency of MABR tank 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:

[0111] 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.

[0112] Data preprocessing: Clean and normalize the collected data to meet the requirements of the LSTM model. Divide the normalized data into training set, validation set and test set to ensure the effectiveness and accuracy of the model optimization process.

[0113] Constructing an LSTM model based on time series to predict the denitrification efficiency of MABR.

[0114]

[0115] where, is the predicted denitrification efficiency, u is the input influent data, which can be represented as U={u x , u Rv , u v}, where u x , u Rv , u v represent the influent water quality, denitrification rate, and influent flow rate, respectively, and d is the bias term.

[0116] Optimize the hyperparameters of the LSTM model using the jumping spider algorithm. The specific process is as follows:

[0117] Initialization: Set the population size, iteration number, and other parameters of the jumping spider algorithm.

[0118] Individual encoding: Encode the hyperparameters of the LSTM model as the position of the jumping spider.

[0119] Search process: The jumping spider algorithm updates the position by simulating the jumping and foraging behavior of spiders to find the optimal parameters.

[0120]

[0121] where S i (t) is the position of the i-th jumping spider at the t-th iteration, ΔS i (t) is the change in position.

[0122] where the change in position is calculated by the following formula:

[0123] ;

[0124] In the formula, S b (t) is the current global optimal position, R i (t) is a random number subject to a certain distribution, used to simulate the jumping behavior of spiders.

[0125] The prediction error of the LSTM model is calculated using the following formula as the fitness of the jumping spider algorithm:

[0126] ;

[0127] In the formula, ey j is the actual MABR tank 2 denitrification efficiency, is the LSTM model predicted denitrification efficiency, and m is the number of samples.

[0128] In each iteration, the following operations are performed according to the fitness: select the solution with better fitness, and update the position of the candidate solution according to its position and fitness. The exploration strategy is introduced, and new solutions are randomly generated within the search space to increase the diversity of the population.

[0129] 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 best hyperparameter configuration.

[0130] Model training and optimization: The LSTM model optimized by the jumping spider algorithm is used to predict the MABR denitrification efficiency. The best hyperparameter setting found is used to train the final LSTM model. The training process is as follows: the training set is used to train the LSTM model, and the loss of the training set and the validation set is observed to prevent overfitting. If the validation loss is found to no longer decrease during training, the training can be stopped early. After each training cycle, the performance of the model is checked using the validation set. The model parameters are adjusted according to the loss of the validation set and the corresponding indicators. During the training process, the best model can be saved, and then this model is used for testing after the model training is completed, and the model parameters that meet the test accuracy are retained.

[0131] 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 clear water flow rate. The optimal denitrification efficiency is predicted by the LSTM model optimized by the jumping spider algorithm, and the corresponding aeration rate is determined, which is expressed by the formula:

[0132] ;

[0133] In the formula, argmax r represents finding the aeration rate that maximizes the predicted denitrification efficiency.

[0134] The jumping spider algorithm can effectively optimize the LSTM model and is suitable for fine-tuning parameters when the optimal solution has been approached. For example, optimization within a specific learning rate range can quickly improve the performance of the LSTM model. By conducting a deep search near the local solution, the jumping spider algorithm can significantly improve the final performance of the model in certain cases, helping to find the best hyperparameters for a specific dataset. The LSTM model optimized by the jumping spider algorithm determines the optimal aeration rate based on the prediction results of the MABR denitrification efficiency to optimize the wastewater denitrification process, achieving the optimal combination of denitrification rate and aeration rate, effectively improving the denitrification efficiency, and further optimizing the water quality.

[0135] The aeration control unit is used to aerate the MABR tank 2 according to the optimal aeration rate.

[0136] In other embodiments, to further improve the denitrification efficiency of the wastewater treatment, the MBR tank 4 can also be subjected to aeration treatment, and the specific aeration rate can also be output by the LSTM model optimized by the jumping spider algorithm.

[0137] Embodiment 2: Please refer to Figure 2 The present embodiment provides a high-efficiency denitrification wastewater treatment method, which can be controlled by the high-efficiency denitrification wastewater treatment system of embodiment 1, comprising the following steps:

[0138] S1: Introduce wastewater into the conditioning tank 1. Add reagents to the conditioning tank 1 and use the sedimentation method for preliminary separation to remove large-particle impurities in the wastewater. The reagent addition method is as follows: detect the wastewater quality data, analyze and predict the water quality data based on the optimized LSTM model. Query the reagent type in the pre-stored control table according to the water quality data, and calculate the reagent amount.

[0139] S2: Deliver the preliminarily separated clean water to the MABR tank 2, use micro-bubble aeration to provide oxygen, and perform nitrification reaction on the wastewater to intercept microorganisms and suspended solids in the wastewater. The aeration rate setting method is as follows: construct and train the LSTM model optimized by the jumping spider algorithm. Output the optimal aeration rate in the LSTM model optimized by the jumping spider algorithm based on the water quality data and the clean water flow rate.

[0140] S3: Deliver the clean water separated by the MABR tank 2 to the AOA tank 3, and sequentially pass through the anaerobic tank 31, the aerobic tank 32, and the anoxic tank 33 for denitrification treatment.

[0141] S4: The MBR tank 4 further purifies the water quality by membrane separation technology and uses the backwashing pump to clean the membrane to maintain the membrane flux.

[0142] S5: Deliver the clean water meeting the water quality standards to the clean water tank 5, and discharge the sludge in the conditioning tank 1, the MABR tank 2, the AOA tank 3, and the MBR tank 4 to the sludge tank.

[0143] Each technical feature of the above-described embodiments can be combined with any other technical feature, and for the sake of brevity, not all possible combinations are described, but it is understood that the scope of the present disclosure encompasses all such possible combinations.

[0144] The above-described embodiments are merely representative of several embodiments of the present disclosure, and the description is relatively specific and detailed, but should not be construed as limiting the scope of the patent. It should be noted that for those skilled in the art, without departing from the concept of the present disclosure, a number of modifications and improvements can be made, which are all within the scope of the present disclosure. Therefore, the scope of the patent of the present disclosure should be subject to the appended claims.

Claims

1. A high efficiency denitrification sewage treatment system, characterized in that, The application relates to a sewage treatment system, which comprises: a conditioning tank for preliminary separation of sewage by sedimentation to remove large-particle impurities in the sewage; a MABR tank for providing oxygen by micro-bubble aeration and carrying out a nitrification reaction on the sewage to intercept microorganisms and suspended matters in the sewage; an AOA tank comprising an anaerobic tank, an aerobic tank and an anoxic tank; the anaerobic tank uses anaerobic microorganisms to ferment organic matters and make sulfates and other oxidants be reduced to form carbon dioxide and ammonia; the aerobic tank uses aerobic microorganisms to oxidize ammonia-nitrogen compounds in the sewage into nitrite and nitrate, and uses denitrifying microorganisms to reduce the nitrate into nitrogen; the anoxic tank uses anaerobic microorganisms to decompose organic matters in the sewage to release ammonia-nitrogen compounds, and converts the ammonia-nitrogen compounds into gaseous nitrogen through a denitrification process, and in an anaerobic environment, uses phosphate bacteria to release phosphorus; an MBR tank for purifying the sewage by a membrane separation technology, and using a backwashing pump to clean an ultrafiltration membrane in the MBR tank to maintain the membrane flux of the ultrafiltration membrane; a conveying module for introducing the sewage into the conditioning tank and conveying the preliminarily separated clean water along the conditioning tank, the MABR tank, the AOA tank and the MBR tank in sequence; a detection module for detecting water quality data of the sewage; a chemical feeder for adding chemicals of corresponding types and weights into the conditioning tank and the MBR tank according to the water quality data; the chemical feeder comprises a chemical proportioning unit and a chemical mixing and conveying unit; the chemical proportioning unit is used for querying chemical types in a pre-stored control table according to clean water quality data, and calculating a chemical amount and a chemical adding rate according to a clean water flow rate data; the chemical mixing and conveying unit is used for mixing chemicals to obtain mixed chemical liquid according to the chemical types and the chemical amount, and conveying the mixed chemical liquid according to the chemical adding rate; the chemical proportioning unit inputs historical water quality data into an LSTM model optimized by a sparrow algorithm to obtain water quality prediction data, and the specific process comprises the following steps: collecting historical water quality data and performing standardization processing; constructing an LSTM model for predicting water quality changes; optimizing hyperparameters of the LSTM model by using the sparrow algorithm; inputting the historical water quality data into the optimized LSTM model to obtain water quality prediction data; an aeration rate regulator for setting an optimal aeration rate according to the water quality data, and adjusting an aeration rate of the MABR tank according to the optimal aeration rate.

2. The high efficiency denitrification wastewater treatment system according to claim 1, wherein The detection module comprises sewage initial data detection unit, separation flow rate detection unit, aeration rate detection unit, water quality detection unit; the sewage initial data detection unit is used for detecting the initial water quality data of sewage; the separation flow rate detection unit is used for detecting the clear water flow rate data, including the rate v1 of outputting clear water from the adjustment tank to the MABR tank, the rate v2 of outputting clear water from the MABR tank to the AOA tank, the rate v3 of outputting clear water from the AOA tank to the MBR tank, and the rate v4 of outputting clear water from the MBR tank to the clear water tank; the aeration rate detection unit is used for monitoring the real-time aeration rate r in the MABR tank; and the water quality detection unit is used for detecting the clear water quality data output by the MABR tank, the AOA tank and the MBR tank.

3. The high efficiency denitrification wastewater treatment system according to claim 1, wherein The specific steps of optimizing the hyperparameters of the LSTM model by using the sparrow algorithm include: encoding the hyperparameters of the LSTM model as the position vector of each sparrow individual; applying the search mechanism of the sparrow algorithm to calculate the new position of each sparrow individual; selecting the sparrow individual with the highest fitness as the leader and the other sparrow individuals as followers; determining whether the termination condition is met, and if yes, outputting the hyperparameters corresponding to the leader with the highest fitness, and if not, updating the positions of the leader and the followers and reselecting the leader according to the fitness of each sparrow individual after the update.

4. The high-efficiency denitrification sewage treatment system according to claim 3, characterized in that, The position update formula of the leader is expressed as: ； wherein is the position of the th sparrow in the th iteration, is the position of the current best solution, is a step size control parameter, is a random number; The position update formula of the follower is expressed as: ； wherein is the position of the current worst solution, is a random number; The calculation formula of the fitness is expressed as: ； In the formula, is the actual value of the water quality data, is the sample number.

5. The high efficiency denitrification wastewater treatment system according to claim 1, wherein The aeration rate regulator comprises 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 for constructing an optimal aeration rate analysis model based on the improved LSTM network of the jumping spider algorithm; the optimal aeration rate calculation unit is used for inputting the water quality data and the clear water flow rate data into the optimal aeration rate analysis model to output the optimal aeration rate; and the aeration control unit is used for aerating the MABR tank according to the optimal aeration rate.

6. The high efficiency denitrification wastewater treatment system according to claim 5, wherein The construction process of the optimal aeration rate analysis model includes: constructing a denitrification simulation database, cleaning and normalizing the denitrification simulation database to obtain preprocessed data; constructing a basic aeration rate analysis model based on the LSTM network; optimizing the hyperparameters of the basic aeration rate analysis model by using the jumping spider algorithm to obtain the optimal aeration rate analysis model; training the optimal aeration rate analysis model by using the preprocessed data and retaining the model parameters that finally meet the test accuracy.

7. The high efficiency denitrification wastewater treatment system according to claim 5, wherein The specific process of optimizing the hyperparameters of the optimal aeration rate analysis model by using the jumping spider algorithm includes: initialization: setting the population size and the number of iterations of the jumping spider algorithm; individual encoding: combining and encoding the hyperparameters of the optimal aeration rate analysis model as the position of the jumping spider; searching process: updating the position of each jumping spider by imitating the jumping and foraging behavior of spiders; the position update formula is expressed as: ； In the formula, is the first jumping spider in the first iteration position, is the change amount of the position; wherein the change of the position is expressed as: ； wherein, is the current global optimal position, is a random number subject to a certain distribution, used to simulate the jumping behavior of spiders; The evaluation process updates the position of the candidate solution according to the position and fitness of each jumping spider; an exploration strategy is introduced to generate a new candidate solution randomly within the search space; the fitness of the jumping spider algorithm is expressed as: ； In the formula, is the actual MABR tank denitrification efficiency, is the denitrification efficiency predicted by the LSTM model, and m is the number of samples. The search process and the evaluation process are cycled until a set termination condition is reached, and the hyperparameter combination corresponding to the jumping spider with the highest fitness is output.

8. A high-efficiency denitrification sewage treatment method, which adopts the high-efficiency denitrification sewage treatment system according to any one of claims 1 to 7, characterized in that, The method comprises the following steps: S1: introducing wastewater into a conditioning tank; Adding reagents to the conditioning tank and using sedimentation method for preliminary separation to remove large particle impurities in the wastewater; wherein the reagent adding method is as follows: detecting wastewater quality data, predicting the water quality data according to the LSTM model optimized by the sparrow algorithm, querying the reagent type in the pre-stored control table according to the water quality data, and calculating the reagent amount; S2: delivering the preliminarily separated clean water to the MABR tank, using micro-bubble aeration to provide oxygen, and carrying out nitrification reaction on the wastewater to intercept microorganisms and suspended solids in the wastewater; wherein the aeration rate setting method is as follows: constructing and training an optimal aeration rate analysis model of the LSTM network optimized by the jumping spider algorithm; according to the water quality data and the clean water flow rate data, outputting the optimal aeration rate in the optimal aeration rate analysis model; S3: delivering the clean water separated by the MABR tank to the AOA tank, and sequentially passing through an anaerobic tank, an aerobic tank and an anoxic tank for denitrification treatment; S4: delivering the clean water separated by the AOA tank to the MBR tank, purifying the water quality by membrane separation technology, and cleaning the ultrafiltration membrane in the MBR tank by using a backwashing pump to maintain the membrane flux of the ultrafiltration membrane; S5: delivering the clean water meeting the water quality standard to a clean water tank, and discharging the sludge in the conditioning tank, the MABR tank, the AOA tank and the MBR tank to a sludge tank.

Images