Method for advanced treatment and reuse of cold-rolled silicon steel magnesium oxide wastewater

Abstract

The application discloses a method for deep treatment and recycling of wastewater of cold-rolled silicon steel and magnesium oxide, and particularly relates to the technical field of wastewater treatment; through pH adjustment, coagulation pretreatment, fine filtration before a membrane and reagent stabilization of the wastewater, stable control of the water quality of the influent is realized; a two-stage reverse osmosis system combined with dynamic flow and a cross-flow structure is adopted, concentration polarization and local stress corrosion risk are significantly inhibited; real-time monitoring of the membrane surface pressure difference and the recovery rate and intelligent algorithms are combined, adaptive optimization and regulation and control of system operation parameters are realized; finally, stable recycling of high-quality produced water and efficient reduction and disposal of concentrated liquid are realized; the application effectively solves the problems of membrane pollution, poor system stability and low recycling efficiency in the prior art, and significantly improves the wastewater resource utilization level and system operation safety.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wastewater treatment, in particular to a method for deep treatment and recycling of magnesium oxide wastewater in cold-rolled silicon steel production. BACKGROUND

[0002] The deep treatment and recycling of magnesium oxide wastewater in cold-rolled silicon steel production refers to the wastewater generated in the production process of cold-rolled silicon steel using magnesium oxide as a neutralizing agent or treatment agent. Through advanced deep treatment technologies such as membrane separation, chemical precipitation, and adsorption, impurities, heavy metals, and other harmful substances in the wastewater are removed, and the treated water quality meets the recycling standards, thereby realizing the recycling of wastewater. In the prior art, reverse osmosis (RO) systems are used to remove dissolved salts and trace heavy metals in water, which is the core equipment to achieve standard recycling water. In a natural state, water molecules will flow from a low concentration to a high concentration through a semi-permeable membrane, which is called "osmosis". In reverse osmosis, by applying external high pressure (higher than the osmotic pressure of the solution), water molecules are forced to flow from a high concentration side (contaminated water) to a low concentration side (pure water), while dissolved salts, heavy metal ions, and organic matter are retained by the membrane.

[0003] The prior art has the following disadvantages:

[0004] During the operation of the reverse osmosis system, if the water flow distribution is uneven or there are defects in the design of the membrane assembly, a concentration polarization zone may be formed on the membrane surface, causing the local area to have an abnormally high solute concentration, exceeding its solubility limit. This local high-concentration environment may cause rapid deposition of salts or formation of chemical stress, leading to local stress corrosion of the membrane material. After long-term operation, "etching points" are formed on the membrane surface, damaging the integrity of the membrane, eventually causing perforation or leakage, which seriously affects the stability of the system and the quality of the effluent. SUMMARY

[0005] The purpose of the present application is to provide a method for deep treatment and recycling of magnesium oxide wastewater in cold-rolled silicon steel production to solve the problems in the background art.

[0006] To achieve the above-mentioned purpose, the present application provides the following technical solution: a method for deep treatment and recycling of magnesium oxide wastewater in cold-rolled silicon steel production, comprising:

[0007] Preliminary treatment of the wastewater containing magnesium oxide in the production process of cold-rolled silicon steel, which includes adjusting the pH, adding coagulants or flocculants;

[0008] Adding anti-fouling agents and membrane protectants to the wastewater, and removing colloids and high-molecular-weight organic matter through a multi-media filter or an ultrafiltration device to adjust the inlet water parameters;

[0009] The wastewater stabilized before the membrane is sequentially fed into at least two reverse osmosis systems for membrane separation treatment, and the reverse osmosis system adopts a dynamic flow regulation and cross-flow distribution structure.

[0010] Real-time monitoring of membrane surface pressure difference and recovery rate parameters, adaptive optimization and control of system operation parameters through intelligent algorithm;

[0011] The reverse osmosis water is collected and reused according to the production process requirements, and the concentrated liquid is sent to the concentration or treatment unit for further disposal.

[0012] Preferably, the dynamic flow regulation includes setting an online flow monitoring and feedback system, and adjusting the water inlet pressure, flow rate and recovery rate of each stage of the reverse osmosis system in real time according to the changes of water quality, conductivity and temperature.

[0013] Preferably, the method for obtaining the membrane surface pressure difference is to install an online pressure sensor to record the instantaneous pressure values of the inlet and outlet of a section of membrane module, and to calculate the membrane surface pressure difference of the section of membrane module: ΔP = P in -P out ; ΔP is the membrane surface pressure difference of a section of membrane module, P in is the instantaneous pressure value of the inlet of the membrane module, and P out is the instantaneous pressure value of the outlet of the membrane module; for a reverse osmosis system composed of multiple sections of membrane shells in series, the pressure difference of each section of membrane shell should be calculated respectively, and then the average pressure difference is summed up as the membrane surface pressure difference, expressed as: In the formula, n is the total number of membrane shells, P in，i and P out，i are the inlet and outlet pressures of the i-th section of membrane shell, and SA is the membrane surface pressure difference.

[0014] Preferably, the method for obtaining the recovery rate is to select the key parameters affecting the recovery rate to form a feature sequence input: Q feed (t) is the water inlet flow rate, Q permeate (t) is the permeate water flow rate, and T(t) is the water temperature; the input is a multivariate sequence within a time window, and the prediction target is the recovery rate prediction value in the future period of time The actual recovery rate R(t) under the interference condition is calculated as: But under the condition of severe fluctuation or sensor error, the LSTM model is used to predict and correct it, realizing SR = LSTM(X t-k+1 ,..., X t ); wherein: X t is all input feature vectors at time t, k is the time step, and SR is the recovery rate.

[0015] Preferably, the intelligent algorithm includes training the historical operation data by using an LSTM neural network to predict the future recovery rate fluctuation trend and optimize the system parameter settings.

[0016] Preferably, the membrane surface pressure difference and the recovery rate are converted into a comprehensive feature vector, the comprehensive feature vector is taken as an input of a machine learning model, the machine learning model takes a membrane pollution risk value label as a prediction target for each set of comprehensive feature vectors, minimizes the sum of prediction errors of all membrane pollution risk value labels as a training target, and trains the machine learning model until the sum of prediction errors converges, and stops model training, and determines the membrane pollution risk value according to the model output result, wherein the machine learning model is a polynomial regression model.

[0017] Preferably, the obtained membrane pollution risk value is compared with a gradient risk threshold value, the gradient risk threshold value includes a first risk threshold value and a second risk threshold value, and the first risk threshold value is less than the second risk threshold value, and the membrane pollution risk value is compared with the first risk threshold value and the second risk threshold value respectively.

[0018] If the membrane pollution risk value is greater than the second risk threshold value, it is determined as a high-risk pollution state, and the system immediately issues an alarm and performs adaptive optimization control on system operation parameters.

[0019] If the membrane pollution risk value is greater than or equal to the first risk threshold value and less than or equal to the second risk threshold value, it is determined as a moderate-risk state, and the system automatically starts low-intensity backwashing and increases the pretreatment load of the membrane.

[0020] If the membrane pollution risk value is less than the first risk threshold value, it is determined as a low-risk operation state, and the system maintains the current operation mode and only records the state for trend analysis.

[0021] Preferably, the concentrated liquid is reduced and discharged or recycled after membrane concentration, evaporation crystallization or chemical neutralization precipitation, distilled water can be reused, and solid residues enter a solid waste treatment system.

[0022] In the above technical solution, the present application provides technical effects and advantages:

[0023] 1. The present application forms a complete and efficient deep purification and recycling method by multi-stage treatment and intelligent control of the cold-rolled silicon steel wastewater containing magnesium oxide.

[0024] 2, The application combines membrane surface pressure difference and recovery rate parameters, builds a membrane pollution risk prediction mechanism through an LSTM neural network and a polynomial regression model, realizes real-time data-based operation state identification and self-adaptive optimization control. The system can automatically adjust the operation strategy according to the membrane pollution risk level, such as pressure adjustment, load reduction, backwashing, etc., thereby prolonging the service life of the membrane module and reducing the frequency of chemical cleaning. The final water can meet the needs of various process water, and the concentrated liquid can be treated by membrane concentration, evaporation or precipitation to realize resource reduction, effectively reducing the environmental pressure. The overall scheme has the advantages of high intelligence, high water resource reuse rate and low operation risk, and is suitable for high-pollution industrial wastewater reuse scenes in steel, metallurgy and other industries. BRIEF DESCRIPTION OF DRAWINGS

[0025] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed in the embodiments will be briefly introduced as follows. Obviously, the drawings described below are only some embodiments described in the present application, and other drawings can also be obtained by those skilled in the art based on these drawings.

[0026] Figure 1 The method flowchart of the present application. DETAILED DESCRIPTION

[0027] In order to make the purpose, technical scheme and advantages of the embodiments of the present application more clear, the technical scheme in the embodiments of the present application will be clearly and completely described below in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0028] Embodiment, please refer to Figure 1 The deep treatment and reuse method of the cold-rolled silicon steel magnesium oxide wastewater described in the present embodiment includes:

[0029] The wastewater containing magnesium oxide in the production process of cold-rolled silicon steel is preliminarily treated, and the treatment includes adjusting pH, adding coagulant or flocculant;

[0030] Anti-fouling agent and membrane protective agent are added to the wastewater, and colloidal and high molecular organic matter is removed by a multi-medium filter or an ultrafiltration device to adjust the water inlet parameters;

[0031] The wastewater stabilized before the membrane is sequentially introduced into at least two reverse osmosis systems for membrane separation treatment, and the reverse osmosis system adopts dynamic flow regulation and cross-flow distribution structure;

[0032] The membrane surface pressure difference and recovery rate parameters are monitored in real time, and the system operation parameters are self-adaptively optimized and controlled through an intelligent algorithm;

[0033] The reverse osmosis product water is collected and reused according to the production process requirements, and the concentrate is sent to the concentration or treatment unit for further disposal.

[0034] In the production of cold-rolled silicon steel, magnesium oxide is often used as a neutralizing agent for protective coating in the annealing process. Subsequently, a large amount of magnesium oxide (MgO), magnesium hydroxide (Mg(OH)2) particles, suspended impurities, part of heavy metal ions (such as Fe 2+ / Fe 3+ ) and oil organic pollutants remain in the subsequent cleaning or annealing wastewater. Therefore, the goal of the preliminary treatment stage is to stabilize the water quality, reduce the pollution load, and ensure the stable operation of the subsequent membrane system.

[0035] pH adjustment: adjust the water body pH to the appropriate range (usually pH 6.5-7.5) to promote the conversion of magnesium oxide and heavy metal ions into a precipitable form.

[0036] Method: If the wastewater is alkaline (pH>9), dilute sulfuric acid, hydrochloric acid or citric acid solution can be added for neutralization. If the wastewater is acidic, appropriate addition of sodium hydroxide or lime milk can be added to adjust to neutral or slightly alkaline.

[0037] Control requirements: Use online pH monitoring and automatic dosing system to accurately control the dosage of the adjusting agent and prevent sharp fluctuations in pH.

[0038] Add coagulant: Form flocculation of small colloidal particles, magnesium oxide particles, organic matter and heavy metals in wastewater.

[0039] Operation control: Set stirring device in the coagulation reaction tank, first fast stirring to promote uniform dispersion of the reagent, then slow stirring to promote flocculation reaction. Control the dosage and reaction time, generally fast stirring time is 1-3 min, slow stirring 10-20 min.

[0040] Precipitation and solid-liquid separation: After flocculation reaction, the wastewater enters the sedimentation tank (can be inclined tube sedimentation tank or vertical flow sedimentation tank), and the flocculation, suspended solids and part of magnesium oxide are removed by gravity. The sludge after sedimentation is periodically removed by sludge removal system and enters the sludge treatment unit. Monitoring parameters: suspended solids concentration (SS), turbidity, magnesium oxide concentration, heavy metal residue.

[0041] After completing the preliminary sedimentation treatment of wastewater, there may still be trace amounts of soluble salts, colloids, organic matter, oil and smaller particle size suspended solids. If these components directly enter the reverse osmosis (RO) system, it will cause membrane surface scaling, pollution, and even irreversible damage. Therefore, before entering the membrane system, the wastewater needs to be further treated for water quality stabilization, including reagent addition and filtration system purification.

[0042] Anti-fouling agent and membrane protection agent: Anti-fouling agent: Inhibit the formation of scale on the membrane surface by dissolved salts in water.

[0043] Common types: Organic phosphorus (such as ATMP, HEDP), sulfonic acid, organic polymer anti-fouling agent.

[0044] Dosage method: Set up an online dosing pump to automatically adjust the dosage according to the concentration of calcium, magnesium, silicon, carbonate and other components in the influent.

[0045] The recommended dosage concentration range is 2-8 mg / L, calculated based on the concentration of concentrated water.

[0046] Key control parameters: LSI (Langmuir saturation index) <0.3 to prevent calcium carbonate scale; silicon concentration controlled below 120 mg / L.

[0047] Membrane protection agent: Inhibit the growth and reproduction of microorganisms on the membrane surface, delay biofouling.

[0048] Optional agents: Non-oxidizing biocides (such as isothiazolinone, DBNPA), slow-release protective agents. Use agents compatible with the membrane material to avoid damaging the membrane structure; if the system is long-term shut down, protective agents need to be added for sealing circulation to maintain the wet state of the membrane system.

[0049] Colloid and organic matter removal:

[0050] Multi-media filter: A layered filter bed usually composed of quartz sand, anthracite, ceramic particles, etc. Function: Remove particulate matter, colloidal impurities, and part of organic matter with particle size >10 μm. Operating parameters: Filter speed controlled at 8-12 m / h; regular backwashing (every 6-12 hours) to restore filter speed.

[0051] Ultrafiltration device: Use hollow fiber or spiral membrane modules with a pore size of 0.01-0.1 μm to further remove small molecule colloids, emulsified oil, organic matter, and part of bacteria. Advantages: High precision of membrane interception, ensuring that the SDI (pollution index) of RO influent is <3; can serve as a high-quality pretreatment barrier for RO, significantly extending the membrane life.

[0052] Influent parameter adjustment, main parameters include: pH control: maintain in the range of 6.5-7.5 to adapt to the requirements of RO membrane; temperature regulation: generally controlled at 20-30℃ to prevent high temperature damage to the membrane; conductivity monitoring: ensure that the conductivity is within the design bearing range of the membrane to avoid overload; flow and pressure: accurately adjust the pump speed and influent pressure through a frequency control system to ensure stable flow rate on the membrane surface and avoid concentration polarization.

[0053] After completing the membrane pre-stabilization treatment, the wastewater enters a set of structurally optimized multi-stage reverse osmosis membrane separation system, including the following specific steps:

[0054] Two-stage series reverse osmosis system is set up:

[0055] First stage reverse osmosis unit (RO-1): receiving pretreated and pre-stabilized wastewater; the main goal is to remove most of the soluble salts, heavy metal ions, organic residues; usually designed for high recovery rate operation (recovery rate is generally 60%~70%); the product water enters the second stage system, and the concentrated water enters the concentrated treatment module.

[0056] Second stage reverse osmosis unit (RO-2): receiving the first stage product water, further improving water quality purity; usually using low pressure operation strategy to reduce energy consumption and control product water conductivity; the product water can be directly reused to the process flow, and the concentrated water can be used as reclaimed water or recharged.

[0057] Dynamic flow regulation system is adopted: online flow monitoring and feedback control system is set up to respond to changes in water quality, pressure and temperature at different time periods; the flow rate and pressure of the first and second stage systems are dynamically adjusted to avoid membrane surface load overload or intensified concentration polarization; the system adjusts the pump speed and pressure valve opening degree based on algorithm model (which can be embedded in PLC or SCADA system).

[0058] Cross-flow distribution structure is set up: each stage of reverse osmosis membrane module adopts cross-flow water inlet design, that is, the water inlet passes through the membrane surface at a certain tangent angle; by adding flow guide device or variable cross-section flow channel structure, the membrane surface water flow disturbance is enhanced to prevent edge zone concentration polarization; multi-stage flow distribution plate is set in the membrane shell to make the water inlet uniformly distributed along the membrane length direction, reducing the formation of dead zones. The advantage is that: compared with straight-flow water inlet, cross-flow design can effectively inhibit membrane surface deposition, improve membrane flux stability, reduce membrane pollution rate, and improve cleaning cycle and overall system operation efficiency.

[0059] Real-time monitoring of membrane surface pressure difference and recovery rate parameters, adaptive optimization control of system operation parameters through intelligent algorithm, specifically:

[0060] Membrane surface pressure difference: reflecting the membrane pollution trend; recovery rate: calculated in real time by water inlet and product water flow meter, reflecting system efficiency; data is transmitted to the central control unit at a frequency of seconds, and trend analysis is carried out in the background.

[0061] Among them, the method for obtaining the membrane surface pressure difference is: online pressure sensor (pressure transmitter) is installed, and the instantaneous pressure values of the inlet (water inlet end) and outlet (concentrated water end) of a section of membrane module are recorded respectively, and the membrane surface pressure difference of a section of membrane module is calculated: ΔP=P in -P out ; ΔP is the membrane surface pressure difference of a section of membrane module, P in is the instantaneous pressure value of the inlet of the membrane module, and P outThe instantaneous pressure value at the outlet of the membrane module; for a reverse osmosis system composed of multiple membrane shells in series (such as a three-stage series), the pressure difference of each membrane shell should be calculated separately, and then the average pressure difference is summarized as the membrane surface pressure difference, expressed as: Where n is the total number of membrane shells, P in，i , P out，i is the inlet and outlet pressure of the i-th membrane shell, and SA is the membrane surface pressure difference.

[0062] In the integrated control system, the membrane surface pressure difference is an important indicator for monitoring water production flow, conductivity, etc. The threshold value can be set through the PLC system, and once the ΔP exceeds the preset range, the pump speed is automatically interlocked and adjusted, the alarm is triggered, or the flushing mode is entered.

[0063] The method for obtaining the recovery rate is to select the key parameters that affect the recovery rate to form a feature sequence input: Q feed (t) is the inlet water flow (changes with time), Q permeate (t) is the permeate water flow and T(t) is the water temperature; the input is a multivariate sequence within a time window, such as the past 60 minutes or the past N sampling points of historical data. The prediction target is the recovery rate prediction value in the future Calculate the actual recovery rate R(t) under the presence of interference: But under the condition of severe fluctuations or sensor errors, the value will be predicted and corrected by the LSTM model, realizing: SR = LSTM(X t-k+1 ,..., X t ); Where: X t is all the input feature vectors at time t, k is the time step, and SR is the recovery rate.

[0064] Where, the LSTM model training steps include: collecting multiple parameter time series data during continuous operation; standardizing input features (such as Min-Max or Z-score); sliding window method to build training set and test set. Build LSTM model structure, training target and loss function, loss function uses mean square error (MSE); optimizer: Adam or RMSprop; use historical operation data to train the model; verify the generalization ability of the model under different water quality and load conditions; if the error is large, add Dropout, batch normalization, more LSTM layers, etc. to adjust.

[0065] The membrane surface pressure difference and the recovery rate are converted into a comprehensive feature vector, the comprehensive feature vector is taken as an input of a machine learning model, the machine learning model takes a membrane pollution risk value label as a prediction target for each set of comprehensive feature vectors, minimizes a sum of prediction errors of all membrane pollution risk value labels as a training target, and trains the machine learning model until the sum of prediction errors converges, and stops model training, and determines the membrane pollution risk value according to a model output result, wherein the machine learning model is a polynomial regression model.

[0066] The obtained membrane pollution risk value is compared with a gradient risk threshold value, the gradient risk threshold value includes a first risk threshold value and a second risk threshold value, and the first risk threshold value is less than the second risk threshold value, and the membrane pollution risk value is compared with the first risk threshold value and the second risk threshold value respectively;

[0067] If the membrane pollution risk value is greater than the second risk threshold value, it is determined as a high-risk pollution state, and the system immediately issues an alarm and performs adaptive optimization control on system operation parameters;

[0068] If the membrane pollution risk value is greater than or equal to the first risk threshold value and less than or equal to the second risk threshold value, it is determined as a moderate-risk state, and the system automatically starts low-intensity backwashing, increases membrane pretreatment load, and improves stability;

[0069] If the membrane pollution risk value is less than the first risk threshold value, it is determined as a low-risk operation state, and the system maintains the current operation mode and only records the state for trend analysis.

[0070] The adaptive optimization control on the system operation parameters is specifically:

[0071] During system operation, key operation parameters are obtained in real time through a sensor network and a data acquisition module, including but not limited to:

[0072] Inlet water flow, permeate water flow, inlet and concentrated water end pressure, membrane surface pressure difference, recovery rate, and water quality indexes such as conductivity, temperature, and pH. All parameters are continuously recorded at a set sampling frequency to form a multi-dimensional time series.

[0073] The system inputs the above data into an analysis module to diagnose the current operation state:

[0074] If the membrane surface pressure difference continuously increases and the recovery rate decreases, it is identified as a membrane pollution trend; if the inlet water flow changes greatly and the produced water is unstable, it is identified as load fluctuation; if the concentrated water pressure abnormally increases, it can be identified as backflow blockage or membrane element pressure loss risk. And whether the subsequent trend is controllable or intervention is needed is judged through a regression model, a prediction algorithm (such as LSTM, linear regression or an empirical curve).

[0075] System set control trigger conditions, such as: recovery rate > 10% / hour; flow, conductivity fluctuation exceeds the set deviation; model prediction results show that the membrane pollution risk value > risk threshold. Meet any condition, that is, into the optimization control process.

[0076] System according to different working conditions and target, dynamic selection control mode, including:

[0077] Increase the water flow or concentrated water reflux flow to improve the cross flow rate, reduce the membrane surface concentration polarization. When the recovery rate is too high to cause pollution, automatically reduce the recovery rate to the recommended value (such as 70% to 65%).

[0078] Adjust the frequency of high pressure pump, control the water pressure, prevent the membrane pressure from exceeding the limit or the pressure difference from being too large. When the pressure difference rises, the system can automatically reduce the pressure to the membrane design safety threshold.

[0079] If the membrane pollution trend is related to abnormal pH, adjust the acid and alkali dosage; if the temperature is too low, reduce the recovery rate or use heating to optimize the water flux.

[0080] Feedback control signal to multiple medium filtration, ultrafiltration and other units to improve the frequency of backwashing; if the SDI value (pollution index) before the membrane rises, trigger the ultrafiltration flushing action.

[0081] The system adjusts the pump speed, valve opening, and reagent dosage rate through automatic control units (such as PLC, frequency converter, automatic valve); all operations are included in the control log to ensure traceability and the possibility of manual intervention.

[0082] After the control is executed, the system re-evaluates the changes in key parameters in a short running period (such as 5 minutes): if the pressure difference decreases and the recovery rate tends to be stable → the control is effective; if the problem persists or worsens → trigger the second strategy or alarm manual intervention. The system records the control action and response effect as data accumulation for subsequent algorithm optimization.

[0083] During the operation of the reverse osmosis system, the water is divided into two parts: product water: clean water treated by RO membrane desalination; concentrated water: high concentration concentrate containing dissolved salt, magnesium oxide residue, heavy metals, etc.

[0084] The recycling of product water includes:

[0085] The product water is measured by an online flowmeter and then enters the product water buffer tank or intermediate storage tank; the storage tank is provided with a liquid level control system, which can automatically supply water according to the water demand or switch to the standby water source.

[0086] Set online conductivity meter, pH meter, thermometer and other equipment to ensure that the product water quality meets the standard; if the fluctuation is abnormal, the system can switch to discharge or bypass pipeline through a three-way valve.

[0087] Possible applications in the cold-rolled silicon steel production process: used as cleaning water after annealing; used as dilution water for magnesium-based coating cleaning or spraying; used to supplement the circulating cooling water system; used as a diluent for descaling agents or pickling processes; used as water for non-contact equipment cooling (such as hydraulic systems).

[0088] If the water hardness and conductivity are too low or too high, the water can be mixed according to the process requirements, such as mixing with fresh water or circulating water for use; a proportional water mixing valve is set to achieve automatic proportioning according to the preset proportion.

[0089] Further disposal of the concentrated liquid: enter the concentration treatment unit, and after the concentrated water is collected, it enters the following possible paths: a. Multi-effect evaporation or mechanical vapor recompression (MVR) system; further concentrate the wastewater volume to recover salt or magnesium ions; distilled water can be returned to the system for reuse, and solid residues enter solid waste treatment. b. Membrane concentration (such as NF / DTRO): applied to concentrated water with relatively large volume and relatively light pollution, to recover some valuable substances such as Mg 2+ ; NF concentrated liquid can be further evaporated or used as a resource material. c. Chemical neutralization + precipitation treatment: add neutralizing agents and flocculants to remove magnesium oxide, heavy metals and other pollutants by sedimentation; the clear liquid is discharged in accordance with the standard, and the sludge is dewatered and sent to a hazardous waste or resource treatment plant for treatment.

[0090] If the concentrated water cannot be treated, it can be discharged to the industrial wastewater pipe network or evaporation pond after meeting the environmental protection standards and being monitored to meet the standards; online monitoring devices (COD, ammonia nitrogen, conductivity, etc.) must be set before discharge to achieve data retention and alarm control.

[0091] The above formulas are dimensionless to calculate their numerical values, and the formulas are obtained by software simulation of a large amount of data to obtain a formula closest to the real situation, and the preset parameters in the formula are set by a person skilled in the art according to the actual situation.

[0092] The above-described embodiments can be implemented in part or in whole through software, hardware, firmware or any combination thereof. When implemented in software, the above-described embodiments can be implemented using one or more computer programs written in any suitable programming language. The computer programs can be stored in one or more computer-readable storage media, such as a memory, a magnetic disk, an optical disk, a hard disk, a floppy disk, a magnetic tape, a memory card, a ROM, a DVD, a Blu-ray Disc, a CD, a semiconductor memory, a flash memory, or the like. The computer programs can be loaded into a computer, a server, a computer network, or the like, and executed. The computer programs can be distributed to computer systems connected to a network, and executed in parallel. The computer programs can be executed by a computer system that is capable of accessing a network, such as the Internet, and executed in parallel.

[0093] It should be understood that the term "and / or" in this document is merely used to describe associated objects, and can represent three conditions: A and / or B, such as A alone, B alone, or A and B together. In addition, the character " / " in this document generally represents an "or" relationship between the front and rear associated objects, but can also represent an "and / or" relationship. Those skilled in the art can realize that the units and algorithm steps of the examples described in combination with the embodiments disclosed in this document can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0094] The above description is merely a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be covered within the protection scope of the present application.

Claims

1. A method for deep treatment and reuse of magnesium oxide wastewater from cold-rolled silicon steel, characterized in that: include: Preliminary treatment of magnesium oxide-containing wastewater from the cold-rolled silicon steel production process includes pH adjustment and addition of coagulants or flocculants. Anti-scaling agents and membrane protectants are added to the wastewater, and colloidal and high molecular weight organic matter are removed by passing it through a multi-media filter or ultrafiltration device, while adjusting the influent parameters; The wastewater, after pre-membrane stabilization, is sequentially fed into at least two stages of reverse osmosis system for membrane separation treatment. The reverse osmosis system adopts a dynamic flow regulation and cross-flow distribution structure. Real-time monitoring of membrane surface pressure difference and recovery rate parameters, and adaptive optimization and control of system operating parameters through intelligent algorithms; The method for obtaining the membrane surface pressure difference is as follows: install an online pressure sensor, record the instantaneous pressure values ​​at the inlet and outlet of a section of the membrane module, and calculate the membrane surface pressure difference of a section of the membrane module. ; The pressure difference across the membrane surface of a membrane module. This represents the instantaneous pressure value at the inlet of the membrane module. This represents the instantaneous pressure value at the membrane module outlet. For a reverse osmosis system consisting of multiple membrane housings connected in series, the pressure difference of each membrane housing should be calculated separately, and then the average pressure difference should be summed as the membrane surface pressure difference. The expression is: In the formula, n is the total number of membrane shells. Let SA be the inlet and outlet pressures of the i-th membrane segment, and SA be the membrane surface pressure difference. The recovery rate is obtained by selecting key parameters that affect the recovery rate to form a feature sequence input: This refers to the inlet water flow rate. Let T(t) be the infiltration flow rate and T(t) be the water temperature; the input is a multivariate series within a time window, and the prediction target is the predicted recovery rate over a future period. ; Calculate the actual recovery rate under interference conditions. : However, under conditions of severe fluctuations or sensor errors, an LSTM model will be used for prediction and correction to achieve [the desired outcome]. ;in: Let k be the total number of input feature vectors at time t, and k be the time step. Recovery rate; The reverse osmosis permeate is collected and reused according to the production process requirements, while the concentrate is sent to a concentration or treatment unit for further processing.

2. The method for deep treatment and reuse of magnesium oxide wastewater from cold-rolled silicon steel according to claim 1, characterized in that: The dynamic flow regulation includes setting up an online flow monitoring and feedback system to adjust the inlet pressure, flow rate and recovery rate of each stage of the reverse osmosis system in real time according to changes in inlet water quality, conductivity and temperature.

3. The method for deep treatment and reuse of magnesium oxide wastewater from cold-rolled silicon steel according to claim 1, characterized in that: The intelligent algorithm includes using an LSTM neural network to train on historical operating data to predict future recovery rate fluctuation trends and optimize system parameter settings.

4. The method for deep treatment and reuse of magnesium oxide wastewater from cold-rolled silicon steel according to claim 3, characterized in that: The membrane pressure difference and recovery rate are converted into a comprehensive feature vector. This comprehensive feature vector is used as the input to a machine learning model. The machine learning model uses the prediction of membrane fouling risk value labels for each set of comprehensive feature vectors as the prediction objective and minimizes the sum of prediction errors for all membrane fouling risk value labels as the training objective. The machine learning model is trained until the sum of prediction errors converges, at which point the model training stops. The membrane fouling risk value is determined based on the model output. The machine learning model is a multinomial regression model.

5. The method for deep treatment and reuse of magnesium oxide wastewater from cold-rolled silicon steel according to claim 4, characterized in that: The obtained membrane fouling risk value is compared with the gradient risk threshold, which includes a first risk threshold and a second risk threshold, and the first risk threshold is less than the second risk threshold. The membrane fouling risk value is compared with the first risk threshold and the second risk threshold respectively. If the membrane fouling risk value is greater than the second risk threshold, it is determined to be a high-risk fouling state, and the system immediately issues an alarm and adaptively optimizes and controls the system operating parameters. If the membrane fouling risk value is greater than or equal to the first risk threshold and less than or equal to the second risk threshold, it is determined to be a medium risk state, and the system will automatically start low-intensity backwashing and increase the membrane pretreatment load. If the membrane fouling risk value is less than the first risk threshold, it is determined to be a low-risk operating state. The system maintains the current operating mode and only records the status for trend analysis.

6. The method for deep treatment and reuse of magnesium oxide wastewater from cold-rolled silicon steel according to claim 5, characterized in that: The concentrated liquid is concentrated through membrane, evaporated and crystallized or chemically neutralized and precipitated to achieve reduced emissions or resource utilization. The distilled water can be reused, and the solid residue enters the solid waste treatment system.

Images