Photo-Fenton coupled fluidized bed reactor and wastewater treatment process

Abstract

The present invention discloses a photo-Fenton coupled fluidized bed reaction device and a wastewater treatment process, which relates to the technical field of wastewater treatment and includes wastewater regulation and reagent pre-dosing; photo-catalytic-fluidized bed coupled reaction, monitoring the flow rate and pressure drop of the fluidized bed reaction, constructing a fluidization identification model and a fluctuation characteristic vector, judging whether the wastewater purification efficiency in the stable fluidization state is in the optimal range, if deviated, excluding interference factors through the causal method of event sequence, after confirming that it is caused by the flow rate, constructing a flow rate control model including a dynamic buffer boundary, and realizing flow rate control by using a closed-loop control and a hierarchical response mechanism; solid-liquid-gas three-phase separation and catalyst recovery, the reaction device includes: a photo-catalytic reaction module, a fluidized bed reactor module, a Fenton reagent dosing module, a three-phase separation module, a circulation system module and an auxiliary equipment module. The present invention is beneficial to achieving a stable wastewater treatment effect through dynamic flow rate control.

Description

Technical Field

[0001] The present invention relates to the technical field of wastewater treatment, and particularly to a photo-Fenton coupled fluidized bed reaction device and a wastewater treatment process. Background Art

[0002] In the field of industrial wastewater treatment, the coupled technology of photocatalysis and Fenton reaction has attracted much attention because it can efficiently generate hydroxyl radicals to degrade organic substances. In the traditional process, in a fluidized bed reactor, the carrier particles loaded with a catalyst are in a fluidized state through the liquid flow, which can strengthen the solid-liquid mass transfer. However, its stable operation depends on precise flow rate control. If the flow rate is too low, the carrier will easily settle and the mass transfer efficiency will decrease; if it is too high, particle erosion and loss will occur, and the energy consumption will increase. However, there are deficiencies in the flow rate control of the existing technology.

[0003] During the wastewater treatment process, fluctuations in parameters such as initial COD, pollutant concentration, and pH will significantly affect the purification efficiency. However, the existing technology lacks a systematic analysis of the causal relationships of multiple variables, and it is easy to misjudge the influence of the flow rate on the efficiency. For example, when the influent pollutant concentration suddenly increases, if its interference is not excluded, the flow rate may be wrongly adjusted, resulting in system oscillation or deterioration of the treatment effect. Summary of the Invention

[0004] The purpose of the present invention is to provide a photo-Fenton coupled fluidized bed reaction device and a wastewater treatment process to solve at least one of the above-mentioned problems of the existing technology.

[0005] The photo-Fenton coupled fluidized bed wastewater treatment process includes the following steps:

[0006] S1. Wastewater adjustment and reagent pre-injection;

[0007] S2. Photocatalysis-fluidized bed coupled reaction and flow rate monitoring;

[0008] S3. Solid-liquid-gas three-phase separation and catalyst recovery;

[0009] S4. Advanced treatment and sludge disposal;

[0010] The method of flow rate monitoring is as follows:

[0011] The fluid flow rate and pressure drop in S2 are collected in real time, and a change curve of the real-time pressure drop and the rising flow rate is established;

[0012] By collecting the flow rate and fluid change characteristics in historical process data, a fluidization identification model is constructed, and the change curve of the real-time pressure drop and the rising flow rate is input into the fluidization identification model to identify the fluidization state of the fluidized bed;

[0013] Based on the determined fluidization state, a fluctuation feature vector is constructed and the sub-states of the fluid state change are identified. If the sub-state is a stable fluidization state, it is identified whether the wastewater purification efficiency in the stable fluidization state is within the optimal range;

[0014] If it is not, it is judged whether the wastewater purification efficiency deviates from the optimal range due to the flow rate. If so, a flow rate control model is constructed to control and adjust the flow rate.

[0015] As a further technical solution of the present invention: the method for constructing the fluidization identification model is as follows:

[0016] Smooth the pressure drop and flow rate in the historical data, identify and process the outliers in the historical data through the 3σ criterion, and construct the change curve of the historical flow rate and pressure drop;

[0017] Verify the change curve of the historical flow rate and pressure drop by Darcy's law, identify the change curve of the historical flow rate and pressure drop through the flow rate demarcation point, and divide the change curve into a laminar flow section and a turbulent flow section;

[0018] Based on the change curves of the historical flow rate and pressure drop in the laminar flow section and the turbulent flow section, construct a fluidization identification model, and verify and correct the fluidization identification model using the Ergun equation.

[0019] As a further technical solution of the present invention: the method for obtaining the flow rate demarcation point is as follows:

[0020] Identify the inflection points of the change curve of the historical flow rate and pressure drop, and divide the change curve of the historical flow rate and pressure drop into two sections based on the inflection points to obtain a low flow rate section and a high flow rate section;

[0021] Perform fitting verification of Darcy's law on the low flow rate section and perform polynomial regression fitting verification on the high flow rate section;

[0022] Analyze the fitting errors of the low flow rate section and the high flow rate section, construct a stage error pair, and select the inflection point with the smallest root mean square error in the stage error pair as the flow rate demarcation point.

[0023] As a further technical solution of the present invention: the method for analyzing the fitting errors of the low flow rate section and the high flow rate section is as follows:

[0024] If there are multiple inflection points in the change curve of the historical flow rate and pressure drop, obtain the root mean square error of the low flow rate stage fitting verification and the root mean square error of the high flow rate stage verification corresponding to each inflection point;

[0025] Construct a stage error pair with the root mean square error of the low flow rate stage fitting verification and the root mean square error of the high flow rate stage verification corresponding to each inflection point.

[0026] As a further technical solution of the present invention: The method for identifying the sub - states of the fluid state change is as follows:

[0027] Based on the determined fluidization state, construct a fluctuation feature vector and identify the sub - states of the fluid state change;

[0028] Based on the random forest algorithm, input the fluctuation feature vector into the random forest algorithm to achieve the identification of the sub - states of the fluidization state.

[0029] As a further technical solution of the present invention: The method for constructing the fluctuation feature vector is as follows:

[0030] Calculate the peak factor, power spectrum entropy, fractal dimension, and Lyapunov exponent of the pressure drop within the monitoring period, and construct a fluctuation feature vector.

[0031] As a further technical solution of the present invention: The method for identifying whether the wastewater purification efficiency in the stable fluidization state is in the optimal interval is as follows:

[0032] Obtain the wastewater purification efficiency characteristics in the stable fluidization state from historical data, and at the same time extract the fluctuation feature vector from the historical data;

[0033] Through the quantile regression algorithm, construct a dynamic interval model with the fluctuation feature vector in the historical data and the wastewater purification efficiency characteristics in the stable fluidization state in the historical data;

[0034] Input the fluctuation feature vector and wastewater purification efficiency of the historical stable fluidization state into the dynamic interval model, and calculate the optimal interval of the wastewater purification efficiency;

[0035] Obtain the wastewater purification efficiency corresponding to the fluctuation feature of the real - time fluidization state, compare it with the wastewater purification efficiency in the optimal interval, and judge whether the wastewater purification efficiency in the stable fluidization state is in the optimal interval.

[0036] As a further technical solution of the present invention: The method for controlling and adjusting the flow rate is as follows:

[0037] If the wastewater purification efficiency deviates from the optimal interval due to the flow rate, determine the optimal range of the flow rate through the optimal interval, and set a dynamic buffer boundary for the optimal range of the flow rate;

[0038] Construct a flow rate control model, input the optimal range of the flow rate and the dynamic buffer boundary into the flow rate control model, and achieve the control and adjustment of the flow rate.

[0039] As a further technical solution of the present invention: The method for judging if the wastewater purification efficiency deviates from the optimal interval due to the flow rate is as follows:

[0040] By means of the event sequence causality method, combined with other interference factors of the wastewater treatment process corresponding to the optimization interval, it is judged whether the deviation of the wastewater purification efficiency from the optimization interval is caused by the flow velocity.

[0041] The photocatalytic-Fenton coupled fluidized bed reactor device includes the following modules:

[0042] Photocatalytic reaction module: Using an ultraviolet mercury lamp as the light source, arranged on the inner wall or central axis of the reactor, electrons and holes are generated by exciting the photocatalyst with light energy, promoting the generation of hydroxyl radicals and driving the Fenton cycle;

[0043] Fluidized bed reactor module: Maintaining the liquid flow rate through a circulation pump, making the filler loaded with the catalyst in a fluidized state, realizing solid-liquid mixing and mass transfer, and the design of the inverted cone hopper and settling area promotes solid-liquid separation;

[0044] Fenton reagent dosing module: Comprising a hydrogen peroxide and ferrous ion dosing unit and a pH adjustment unit, creating an acidic condition for the Fenton reaction;

[0045] Three-phase separation module: Separating the treated water, reaction gas and carrier particles through a lamella and a collecting tank, realizing efficient three-phase separation of solid, liquid and gas;

[0046] Circulation system module: Controlling the liquid reflux ratio to promote uniform mixing of the reagent and the wastewater;

[0047] Auxiliary equipment module: Including an adjustment tank, a sedimentation tank and a sludge treatment unit, assisting the main device to operate stably and treating by-products.

[0048] Advantages of the present invention:

[0049] 1. By filling a quartz sand carrier loaded with a titanium dioxide photocatalyst in the fluidized bed reactor and combining with ultraviolet mercury lamp irradiation, electrons and holes are generated by exciting the photocatalyst, and strong oxidizing hydroxyl radicals are generated synergistically with the Fenton reaction, realizing the deep oxidation and degradation of organic matter. The coupling mechanism of photocatalysis and Fenton reaction not only drives the Fe 2+ / Fe 3+ cycle, reducing the generation of iron sludge, but also adsorbing Fe 2+ on the γ-FeOOH catalytic film on the surface of the carrier and promoting the decomposition of hydrogen peroxide, improving the catalyst utilization rate and reaction sustainability, and significantly enhancing the wastewater treatment capacity.

[0050] [[ID=XXX]]2. By collecting the fluid flow velocity and pressure drop in real time, combined with Darcy's law and the Ergun equation, a fluidization identification model is constructed to divide the fixed bed, fluidized bed and transport bed states; based on classical fluid mechanics theory, scientific classification of the fluidization state is realized, the flow characteristics in different flow velocity intervals are clarified, providing a theoretical basis for flow velocity control, and reducing the possibility of process failure caused by misjudgment of the flow regime.

[0051] 3. Construct a fluctuation feature vector using fractal and chaotic features such as the peak factor and power spectrum entropy. Split the sub-states through the random forest algorithm, and dynamically define the optimal interval of the wastewater purification efficiency based on quantile regression. Capture the complex dynamic features of the pressure drop signal through fractal and chaotic theories, and combine the ensemble learning algorithm to realize the classification of the fluidization state, which can identify the subtle state differences that are difficult to distinguish by traditional single parameters, and provide a state criterion for the dynamic evaluation of the wastewater purification efficiency.

[0052] 4. When the efficiency deviates, exclude interference factors such as the initial COD and pollutant concentration through the event sequence causality method. After confirming that it is a flow rate problem, set a dynamic buffer boundary, and adopt a hierarchical response mechanism combined with a closed-loop control model to realize flow rate control, reduce the possibility of mass transfer failure or carrier loss caused by improper flow rate, and maintain the optimal degradation efficiency under a stable fluidization state. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0054] Figure 1 is a flowchart of the wastewater treatment process provided in Embodiment 1 of the present invention;

[0055] Figure 2 is a flowchart of a liquid flow rate monitoring method provided by the present invention;

[0056] Figure 3 is a module diagram of a liquid flow rate monitoring system provided in Embodiment 6 of the present invention;

[0057] Figure 4 is a module diagram of the photo-Fenton coupled fluidized bed reactor provided in Embodiment 7 of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0058] In order to enable those skilled in the art to better understand the solution of the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0059] Embodiment 1

[0060] AsFigure 1 As shown in the figure, the wastewater treatment process provided by the embodiment of the present invention specifically includes the following steps:

[0061] S1. Wastewater regulation and preliminary chemical dosing;

[0062] Industrial wastewater enters the regulating tank of the auxiliary equipment module, and the water quality and quantity are balanced through mechanical stirring;

[0063] The pH monitoring and adjustment system built in the regulating tank preliminarily adjusts the pH of the wastewater to 5, reducing the chemical consumption for subsequent acidic condition adjustment;

[0064] The wastewater enters the Fenton reagent dosing module, and hydrogen peroxide and ferrous ion solution are dosed through a metering pump. At the same time, sulfuric acid is dosed to adjust the pH of the reaction solution system to 3, creating a suitable acidic environment for subsequent reactions;

[0065] S2. Photocatalysis-fluidized bed coupling reaction and flow rate monitoring;

[0066] The pretreated wastewater enters from the inverted cone hopper at the bottom of the fluidized bed reactor and comes into full contact with the quartz sand carrier loaded with titanium dioxide photocatalyst in the reactor;

[0067] The circulation pump of the circulation system module drives the wastewater to flow back from the top to the bottom, maintaining the liquid upward flow rate in the fluidized bed at 60 m / h, making the carrier in a fluidized state, ensuring uniform mixing and efficient mass transfer of the solid-liquid two phases;

[0068] The ultraviolet mercury lamp arranged on the inner wall or central axis of the reactor emits ultraviolet light of 254 nm, exciting the titanium dioxide on the surface of the carrier to generate electron-hole pairs. The holes oxidize hydrogen peroxide to generate hydroxyl radicals (•OH), and the electrons reduce Fe 3+ to Fe 2 + , driving the Fenton reaction to achieve deep oxidation and degradation of organic matter. The residence time of the Fenton reaction is controlled at 1.5 hours;

[0069] It should be explained that the Fenton reaction is Fe 2+ +H2O2→•OH+Fe 3+ ;

[0070] S3. Solid-liquid-gas three-phase separation and catalyst recovery;

[0071] The mixed liquid after the S2 reaction rises to the three-phase separation area of the three-phase separation module at the top of the reactor. Through the gravity sedimentation of the inclined plate assembly, the carrier particles loaded with the catalyst fall back to the downflow area and return to the packing layer with the circulating reflux liquid for reuse;

[0072] The treated supernatant is collected and discharged through the water collection tank, and the gas generated by the reaction is released from the top exhaust port;

[0073] It should be noted that the gases generated by the reaction include CO2, O2, etc.; during the three-phase separation of solid, liquid and gas and the catalyst recovery process, the γ-FeOOH catalytic film formed on the carrier surface adsorbs Fe in water 2+ and catalyzes the decomposition of H2O2, and the photocatalytic synergy promotes the Fe 3+ reduction cycle, reducing the formation of iron sludge and improving the catalyst utilization rate;

[0074] S4, advanced treatment and sludge disposal;

[0075] The effluent after S3 separation enters the sedimentation tank of the auxiliary equipment module, and PAC / PAM is added for coagulation sedimentation to further remove residual iron ions and suspended solids. At the same time, NaOH is added to adjust the pH to 7, so that the total iron concentration of the effluent is 4.0 mg / L;

[0076] It should be noted that polyaluminum chloride (abbreviated as PAC), also known as basic aluminum chloride or hydroxyaluminum chloride, can rapidly form precipitates of colloids in sewage or sludge through it or its hydrolysis products, facilitating the separation of large particle precipitates;

[0077] Polyacrylamide (abbreviated as PAM), commonly known as flocculant or coagulant, is used to treat fine particles and colloidal substances;

[0078] The sludge generated by sedimentation is discharged into the thickening tank, dehydrated by a plate and frame filter press and then transported out for disposal. The filtrate is refluxed to the adjustment tank for cyclic treatment, realizing sludge reduction and recycling of water resources.

[0079] Example 2

[0080] As Figure 1 shown, the wastewater treatment process provided by the embodiment of the present invention specifically includes the following steps:

[0081] S1, wastewater adjustment and chemical pre-injection;

[0082] Industrial wastewater enters the adjustment tank of the auxiliary equipment module, and the water quality and quantity are balanced by mechanical stirring;

[0083] The pH monitoring and adjustment system built in the adjustment tank initially adjusts the pH of the wastewater to 4, reducing the chemical consumption for subsequent acidic condition adjustment;

[0084] The wastewater enters the Fenton reagent injection module, and hydrogen peroxide and ferrous ion solutions are injected through a metering pump. At the same time, sulfuric acid is added to adjust the pH of the reaction solution system to 2.5, creating a suitable acidic environment for subsequent reactions;

[0085] S2, photocatalysis-fluidized bed coupling reaction;

[0086] The pretreated wastewater enters from the inverted cone hopper at the bottom of the fluidized bed reactor and comes into full contact with the quartz sand carrier loaded with titanium dioxide photocatalyst filled in the reactor;

[0087] The circulation pump of the circulation system module drives the wastewater to flow back from the top to the bottom, maintaining the upward liquid flow velocity in the fluidized bed at 30, making the carrier in a fluidized state, ensuring uniform mixing of the solid-liquid two-phase and efficient mass transfer;

[0088] The ultraviolet mercury lamp arranged on the inner wall or central axis of the reactor emits ultraviolet light of 254 nm, exciting the titanium dioxide on the surface of the carrier to generate electron-hole pairs. The holes oxidize hydrogen peroxide to generate hydroxyl radicals (•OH), and the electrons reduce Fe 3+ to Fe 2 + , driving the Fenton reaction to achieve deep oxidation and degradation of organic matter. The residence time of the Fenton reaction is controlled within 1 hour;

[0089] It should be noted that the Fenton reaction is Fe 2+ +H2O2→•OH+Fe 3+ ;

[0090] S3, Solid-liquid-gas three-phase separation and catalyst recovery;

[0091] The mixed liquid after the S2 reaction rises to the three-phase separation area of the three-phase separation module at the top of the reactor. Through the gravity sedimentation of the inclined plate assembly, the carrier particles loaded with the catalyst fall back to the downflow area and return to the packing layer with the circulating reflux liquid for reuse;

[0092] The treated supernatant is collected and discharged through the water collection tank, and the gas generated by the reaction is released from the top exhaust port;

[0093] It should be noted that the gas generated by the reaction includes CO2, O2, etc.; during the solid-liquid-gas three-phase separation and catalyst recovery process, the γ-FeOOH catalytic film formed on the surface of the carrier adsorbs Fe 2+ in the water and catalyzes the decomposition of H2O2. The photocatalytic synergy promotes the Fe 3+ reduction cycle, reducing the generation of iron sludge and improving the catalyst utilization rate;

[0094] S4, Advanced treatment and sludge disposal;

[0095] The effluent after S3 separation enters the sedimentation tank of the auxiliary equipment module, and PAC / PAM is added for coagulation sedimentation to further remove residual iron ions and suspended solids. At the same time, NaOH is added to adjust the pH to neutral 7.5, making the total iron concentration of the effluent 4.5 mg / L;

[0096] The sludge produced by precipitation is discharged into the thickening tank, dehydrated by a plate and frame filter press, and then transported out for disposal. The filtrate is refluxed to the regulation tank for cyclic treatment, realizing sludge reduction and recycling of water resources.

[0097] Example 3

[0098] As Figure 1 shown, the wastewater treatment process provided by the embodiment of the present invention specifically includes the following steps:

[0099] S1. Wastewater regulation and preliminary chemical agent dosing;

[0100] Industrial wastewater enters the regulation tank of the auxiliary equipment module, and the water quality and quantity are balanced through mechanical stirring;

[0101] The pH monitoring and adjustment system built in the regulation tank initially adjusts the pH of the wastewater to 6, reducing the consumption of chemical agents for subsequent acidic condition adjustment;

[0102] The wastewater enters the Fenton reagent dosing module, and hydrogen peroxide and ferrous ion solutions are dosed through a metering pump. At the same time, sulfuric acid is dosed to adjust the pH of the reaction solution system to 4, creating a suitable acidic environment for subsequent reactions;

[0103] S2. Photocatalysis-fluidized bed coupling reaction;

[0104] The pretreated wastewater enters from the inverted cone hopper at the bottom of the fluidized bed reactor and is in full contact with the quartz sand carrier loaded with titanium dioxide photocatalyst in the reactor;

[0105] The circulation pump of the circulation system module drives the wastewater to flow back from the top to the bottom, maintaining the upward flow velocity of the liquid in the fluidized bed at 100 m / h, making the carrier in a fluidized state, ensuring uniform mixing and efficient mass transfer of the solid-liquid two phases;

[0106] The ultraviolet mercury lamp arranged on the inner wall or central axis of the reactor emits ultraviolet light of 254 nm, exciting the titanium dioxide on the surface of the carrier to generate electron-hole pairs. The holes oxidize hydrogen peroxide to generate hydroxyl radicals (•OH), and the electrons reduce Fe 3+ to Fe 2 + , driving the Fenton reaction to achieve deep oxidation and degradation of organic matter. The residence time of the Fenton reaction is controlled within 2 hours;

[0107] It should be noted that the Fenton reaction is Fe 2+ +H2O2→•OH+Fe 3+ ;

[0108] S3. Solid-liquid-gas three-phase separation and catalyst recovery;

[0109] The mixed solution after the S2 reaction rises to the three-phase separation area of the three-phase separation module at the top of the reactor. Through the gravity sedimentation of the inclined plate assembly, the carrier particles loaded with the catalyst fall back to the downflow area and return to the packing layer with the circulating reflux liquid for reuse;

[0110] The treated supernatant is collected and discharged through the water collection tank, and the gas generated by the reaction is released from the top exhaust port;

[0111] It should be explained that the gas generated by the reaction includes CO2, O2, etc.; during the three-phase separation of solid, liquid and gas and the catalyst recovery process, the γ-FeOOH catalytic film formed on the surface of the carrier adsorbs Fe in water 2+ and catalyzes the decomposition of H2O2, and the photocatalytic synergistic effect promotes the Fe 3+ reduction cycle, reducing the generation of iron sludge and improving the utilization rate of the catalyst;

[0112] S4, advanced treatment and sludge disposal;

[0113] The effluent after the S3 separation enters the sedimentation tank of the auxiliary equipment module, and PAC / PAM is added for coagulation sedimentation to further remove residual iron ions and suspended solids. At the same time, NaOH is added to adjust the pH to neutral 7.8, so that the total iron concentration of the effluent is 5 mg / L;

[0114] The sludge generated by the sedimentation is discharged into the thickening tank, dehydrated by a plate and frame filter press and then transported for disposal. The filtrate is refluxed to the regulation tank for circular treatment, realizing sludge reduction and water resource recycling.

[0115] Example 4

[0116] In the photo-Fenton coupled fluidized bed wastewater treatment process, the flow rate is the core operating parameter that determines the fluidization state of the carriers in the fluidized bed, directly affecting the solid-liquid mass transfer efficiency, catalyst stability and reaction system energy consumption. Specifically:

[0117] The fluidized bed reactor makes the quartz sand carriers loaded with the catalyst in a fluidized state through the liquid flow rate. If the flow rate is too low, the carriers will settle (fixed bed state), resulting in insufficient mass transfer area and incomplete reaction; if the flow rate is too high, excessive erosion of the carriers (transport bed state) will be caused, resulting in catalyst loss and a sharp increase in energy consumption. By real-time monitoring the flow rate and combining data such as pressure drop, identifying the current fluidization state (fixed bed / fluidized bed / transport bed), the reactor is in a stable fluidization state with high mass transfer, reducing the possibility of the purification efficiency decline or equipment operation risks caused by improper flow rate.

[0118] As Figure 2 shown, the wastewater treatment process of the photo-Fenton coupled fluidized bed reaction also includes a liquid flow rate monitoring method to realize the monitoring and control of the liquid upward flow rate in the S2 step, including the following steps;

[0119] Step 1: Collect the fluid flow rate and pressure drop in S2 in real time, and establish a change curve of the real-time pressure drop and the upward flow rate.

[0120] Preferably, an electromagnetic flowmeter is installed in the fluidized bed reactor in S2 to collect the upward flow rate of the fluid in the main fluidized bed in real time.

[0121] Pressure sensors are installed at the inlet and outlet of the fluidized bed reactor to measure the pressure drop between the inlet and outlet of the fluidized bed reactor.

[0122] Based on the upward flow rate of the fluid and the pressure drop of the packing layer, and perform smoothing processing on the pressure drop and flow rate, and construct a change curve of the real-time flow rate and pressure drop.

[0123] Step 2: By collecting the flow rate and fluid change characteristics in the historical process data, construct a fluidization identification model, and input the change curve of the real-time pressure drop and upward flow rate into the fluidization identification model to identify the fluidization state of the fluidized bed.

[0124] Obtain the upward flow rate and pressure drop of the photo-Fenton coupled fluidized bed reaction from the historical process data, perform smoothing processing on the pressure drop and flow rate in the historical data, identify and process the outliers in the historical data through the 3σ criterion, and construct a change curve of the historical flow rate and pressure drop.

[0125] Verify the change curve of the historical flow rate and pressure drop by Darcy's law, and identify the change curve of the historical flow rate and pressure drop as the laminar flow section and the turbulent flow section through the flow rate demarcation point.

[0126] Among them, the method for verifying Darcy's law is:

[0127] Identify the inflection point of the change curve of the historical flow rate and pressure drop, and divide the change curve of the historical flow rate and pressure drop into two sections based on the inflection point to obtain the low flow rate section and the high flow rate section.

[0128] Perform fitting verification of Darcy's law on the low flow rate section, and the fitting model is: ;

[0129] Perform polynomial regression fitting verification on the high flow rate section, and the fitting model is: ;

[0130] Among them, is the pressure drop, is the fluid flow rate, and a, b, c, d, and e are the coefficients of the fitting model respectively.

[0131] It should be noted that in the present invention, the fitting model coefficients (a, b, c, d, e) of the low-flow-rate section and the high-flow-rate section are determined by the least squares method: for the low-flow-rate data, the Darcy's law linear model is adopted, and by minimizing the sum of the squares of the residuals between the measured pressure drop and the model prediction values, the linear coefficient b reflecting the viscous resistance and the fixed resistance intercept a are solved; for the high-flow-rate data, a quadratic polynomial model is adopted. Based on the characteristic that the turbulent inertial resistance dominates, the coefficients including the turbulent inertial resistance (e), the transitional zone mixed resistance (d), and the comprehensive fixed resistance (c) are determined by minimizing the sum of the squares of the residuals through multiple linear regression. During implementation, abnormal data are first removed by the 3σ criterion, and the historical flow rate and pressure drop data are smoothed. Then, the inflection points are identified to divide the sections, ensuring that the model fits the laminar-turbulent flow characteristics and providing a basis for accurate identification of the fluidization state;

[0132] If there are multiple inflection points in the change curve of the historical flow rate and pressure drop, obtain the root mean square error of the low-flow-rate stage fitting verification and the root mean square error of the high-flow-rate stage verification corresponding to each inflection point;

[0133] Construct a stage error pair with the root mean square error of the low-flow-rate stage fitting verification and the root mean square error of the high-flow-rate stage fitting verification corresponding to each inflection point;

[0134] Obtain the stage error pairs of all inflection points, and select the inflection point with the smallest root mean square error in the stage error pairs as the flow rate demarcation point;

[0135] It should be noted that the inflection point with the smallest root mean square error in the stage error pair refers to the smallest comprehensive error of the root mean square error of the low-flow-rate stage fitting verification and the root mean square error of the high-flow-rate stage verification;

[0136] The comprehensive error can be obtained by weighted calculation of the root mean square error of the low-flow-rate stage fitting verification and the root mean square error of the high-flow-rate stage verification;

[0137] Based on the flow rate demarcation point, divide the change curve of the historical flow rate and pressure drop into a laminar flow section and a turbulent flow section;

[0138] Based on the change curves of the historical flow rate and pressure drop in the laminar flow section and the turbulent flow section, construct a fluidization identification model, and use the Ergun equation to verify and correct the fluidization identification model;

[0139] Among them, the Ergun equation is: , L is the bed height, dl is the particle diameter in the fluidized bed, n is the fluid viscosity, is the fluid density, is the fluid flow rate, is the bed voidage;

[0140] It should be noted that the Ergun equation is a classical equation that describes the relationship between the pressure drop of fluid flow and the flow velocity in a fixed-bed reactor. The Ergun equation can be used to verify the accuracy of the model. The Ergun equation comprehensively considers the effects of laminar flow and turbulent flow, and can more comprehensively describe the flow characteristics of fluid in a granular bed layer;

[0141] The method of using the Ergun equation to correct the fluidization identification model constructed based on the laminar flow section and the turbulent flow section is as follows: Match the parameters of the Darcy's law model (laminar flow region) and the polynomial model (turbulent flow region) obtained by piecewise fitting with the theoretical form of the Ergun equation. Extract the coefficients corresponding to the laminar term (linear term) and the turbulent term (quadratic term) in the Ergun equation as theoretical constraints. Compare the fitting coefficients of the existing model (such as b of the Darcy model, d and e of the polynomial model). By minimizing the root mean square error between the model prediction value and the calculated value of the Ergun equation, adjust the coefficients and the position of the demarcation point of the piecewise model to ensure the pressure drop continuity and physical consistency of the model in the laminar-turbulent transition region, and finally form a corrected fluidization identification model that conforms to the principles of fluid mechanics;

[0142] Obtain the change curves of the real-time pressure drop and the upward flow velocity, input them into the fluidization identification model, and identify the fluidization state of the fluidized bed;

[0143] It should be noted that by collecting the upward flow velocity and the pressure drop of the fluid in the fluidized bed in real time and constructing a change curve, after smoothing processing and removing outliers by the 3σ criterion, it is matched with the fluidization identification model constructed by fitting with historical data: First, based on Darcy's law and the Ergun equation, divide the low-flow-velocity laminar flow section (suitable for linear fitting of Darcy's law) and the high-flow-velocity turbulent flow section (suitable for polynomial regression fitting) by identifying the inflection point of the curve. Combine the Ergun equation to correct the model parameters to conform to the principles of fluid mechanics. Then, according to the distribution characteristics of the real-time curve in the laminar-turbulent section, judge whether the fluidized bed is in the fixed bed state (the flow velocity is lower than the minimum fluidization velocity, and the pressure drop increases linearly), the fluidized bed state (the flow velocity is between the minimum fluidization velocity and the terminal velocity, and the particles are suspended and fluidized), or the transport bed state (the flow velocity exceeds the terminal velocity, and the particles are entrained and transported), so as to realize the identification of the fluidization state;

[0144] Those skilled in the art can understand that the fluidization state of the fluidized bed is divided into the fixed bed stage, the fluidized bed stage, and the output bed stage;

[0145] When the fluid flow velocity is lower than the minimum fluidization velocity, the bed layer is in the fixed bed stage. At this time, the particles are closely packed and stationary, and only the fluid flows in the narrow channels between the particles. The fluid flow is mainly laminar, and the pressure drop increases linearly with the flow velocity, which conforms to Darcy's law. Its essence is the result of the fluid overcoming the viscous resistance on the particle surface. The bed layer height remains in the initial packing state, there is no macroscopic relative movement between the particles, and the whole shows "solid" characteristics;

[0146] When the flow velocity exceeds the minimum fluidization velocity, the fluid drag force is sufficient to lift the particles, and the bed enters the fluidized bed stage. At this time, the particles break away from the packed state, exhibit suspended and random motion, the bed expands and the height increases with the increase of the flow velocity, showing fluidity similar to that of a liquid;

[0147] When the flow velocity further increases to exceed the terminal velocity of the particles, the bed enters the transport bed stage. At this time, the fluid drag force completely overcomes the particle gravity, the particles are entrained by the fluid and continuously transported, the bed structure is destroyed, and it transforms into a gas-solid / liquid-solid two-phase flow transport state. The pressure drop rises significantly again with the increase of the flow velocity, mainly due to the acceleration of the particles and the collision resistance during the transport process. The flow pattern is mainly turbulent, and the particle concentration gradually decreases along the flow direction.

[0148] It should be explained that the role of identifying the fluidization state of the fluidized bed is as follows:

[0149] Function 1. Ensure the solid-liquid mass transfer efficiency: clarify whether the current flow regime is in the fluidized bed stage of efficient mass transfer, avoid the inefficient or abnormal states of the fixed bed or transport bed, and ensure uniform fluidization of the carrier and full mixing of solid and liquid;

[0150] Function 2. Support precise flow velocity regulation: by identifying the inflection point of the laminar-turbulent transition zone, divide the low-flow-velocity laminar zone and the high-flow-velocity turbulent zone, and combine the fitting model to quantify the relationship between the flow velocity and the pressure drop, providing a theoretical boundary for the closed-loop control model to achieve the adjustment of the flow velocity within the optimal range;

[0151] Function 3. Maintain the stable operation of the system: combine the fluctuation eigenvector and the random forest algorithm to split the sub-states, give real-time warnings of abnormal flow regimes, reduce the possibility of efficiency fluctuations or equipment damage caused by misjudgment of the flow regime, and achieve efficient coupling of photocatalysis and Fenton reaction under a stable fluidized state, improving the robustness and reliability of the wastewater treatment process.

[0152] The technical solution of this embodiment is: collect the fluid flow velocity and pressure drop in S2 in real time, and establish a change curve of the real-time pressure drop and the rising flow velocity; by collecting the flow velocity and fluid change characteristics in the historical process data, construct a fluidization identification model, input the change curve of the real-time pressure drop and the rising flow velocity into the fluidization identification model, identify the fluidization state of the fluidized bed, clarify the flow characteristics in different flow velocity intervals, provide a theoretical basis for flow velocity control, and reduce the possibility of process failure caused by misjudgment of the flow regime.

[0153] Embodiment 5

[0154] As Figure 2 shown, a method for monitoring the liquid flow velocity further includes the following steps:

[0155] Step 3: Based on the determined fluidization state, construct a fluctuation feature vector and identify the sub-states of the fluid state change. If the sub-state is a stable fluidization state, identify whether the wastewater purification efficiency in the stable fluidization state is within the optimized range;

[0156] Based on the determined fluidization state, construct a fluctuation feature vector and identify the sub-states of the fluid state change;

[0157] Preferably, the sub-states include a stable fluidization state, a chaotic fluidization state, and a critical fluidization state;

[0158] Among them, the construction method of the fluctuation feature vector is as follows:

[0159] Calculate the peak factor CF, power spectral entropy PSE, fractal dimension FD, and Lyapunov exponent LE of the pressure drop within the monitoring period;

[0160] Preferably, the acquisition method of the peak factor CF is as follows:

[0161] Through the formula: Obtain the peak factor CF, where is the pressure drop collected in real time by the pressure sensor within the monitoring period, are the maximum pressure drop and the minimum pressure drop within the monitoring period, respectively, is the average pressure drop within the monitoring period;

[0162] Preferably, the acquisition method of the power spectral entropy PSE is as follows:

[0163] Through the formula: Obtain the power spectral entropy PSE;

[0164] Among them, the acquisition method of PS i is as follows:

[0165] Perform a fast Fourier transform on the pressure drop in the monitoring period to convert the pressure drop in the time domain into the pressure drop in the frequency domain, and obtain the discrete frequency component f i corresponding power S(f i );

[0166] Calculate the proportion of the power of each discrete frequency component in the total power to obtain PS i , where i is the number of the frequency component and n is the total number of frequency components;

[0167] Use the box dimension method to calculate the fractal dimension of the pressure drop signal to obtain the fractal dimension FD;

[0168] Calculate the maximum Lyapunov exponent of the pressure drop signal as the Lyapunov exponent LE;

[0169] Those skilled in the art can understand that when calculating the fractal dimension FD of the pressure drop signal using the box dimension method, the pressure drop signal is regarded as a curve in a two-dimensional space, and a square box with a side length of is used to cover the curve, and the number of boxes required for complete coverage is counted. Continuously change and repeat the counting. Through and 's fitting relationship curve, the absolute value of the slope of the fitting relationship curve is the fractal dimension FD;

[0170] The fractal dimension FD is used to represent the complexity and irregularity of the pressure drop signal;

[0171] Perform phase space reconstruction on the signal, convert the time series into points in a multi-dimensional phase space, find the nearest neighbor points of each point in the phase space, track the distance changes of these points as they evolve over time, and calculate the exponential growth rate. The largest exponential growth rate is the maximum Lyapunov exponent LE, and LE can reflect the chaotic degree and dynamic evolution characteristics of the pressure drop signal;

[0172] Based on the peak factor CF, power spectrum entropy PSE, fractal dimension FD, and Lyapunov exponent LE as fluctuation characteristics, construct a fluctuation feature vector ;

[0173] Based on the random forest algorithm, input the fluctuation feature vector into the random forest algorithm to realize the identification of sub-states of the fluidization state;

[0174] Those skilled in the art can understand that the random forest algorithm, as an ensemble learning method, can achieve the splitting of sub-states of the fluidization state by constructing multiple decision trees and using the Bagging strategy for training, and can effectively capture complex non-linear relationships and patterns in the data;

[0175] Use the fluctuation feature vector as the input to train the random forest model. Each decision tree randomly samples the features and samples, optimizes the learning of the mapping relationship from features to sub-states through node splitting, and reduces the risk of overfitting; in the prediction stage, after the new fluctuation feature vector is input, multiple decision trees independently predict and determine the final sub-state through a voting mechanism; the "ensemble decision" method can not only detect the subtle differences of different sub-states in the feature space (such as low CF and PSE in the stable fluidization state, high FD and LE in the chaotic fluidization state), enhance the robustness of the model to noise and data fluctuations, and thus achieve the splitting of sub-states of the fluidization state;

[0176] It should be noted that the parameter combination of the number of trees and the maximum depth in the random forest algorithm is determined through grid search cross-validation. By defining the parameter search space, for example, the number of trees is [50, 100, 150, 200], and the maximum depth is [None, 5, 10, 15]. Then, the labeled historical fluctuation feature vector dataset is divided into a training set and a validation set using k-fold cross-validation. GridSearchCV is used to traverse all parameter combinations, and the weighted F1 score of sub-state classification (as the evaluation metric) is used to calculate the average score of each set of parameters on each validation set. The number of trees and the maximum depth in the parameter combination that maximizes the weighted F1 score are selected as the optimal parameters;

[0177] If the sub-state is the stable fluidization state, the method for identifying whether the wastewater purification efficiency in the stable fluidization state is in the optimal interval is as follows:

[0178] Obtain the wastewater purification efficiency characteristics in the stable fluidization state from historical data, including the COD removal rate and the pollutant degradation rate;

[0179] At the same time, extract the fluctuation feature vectors from the historical data, and use the quantile regression algorithm to construct a dynamic interval model for the fluctuation feature vectors in the historical data and the wastewater purification efficiency characteristics in the stable fluidization state of the historical data;

[0180] Through the formula: , establish a quantile model , where X is the fluctuation feature vector in the historical data, is the wastewater purification efficiency characteristic, is the quantile, corresponding regression coefficient vector for the quantile , obtained through quantile regression training on historical data;

[0181] It should be noted that quantile regression can fit models for different quantiles to characterize the distribution of data at specific quantiles. For the wastewater purification efficiency in the stable fluidization state, it is affected by fluctuation feature vectors (CF, PSE, FD, LE, etc.), and the data may not follow a normal distribution. Quantile regression does not rely on data distribution assumptions and can flexibly capture the non-linear relationship between efficiency and fluctuation features, providing more practical boundaries for constructing the optimal interval;

[0182] Input the fluctuation feature vectors and wastewater purification efficiency of the historical stable fluidization state into the dynamic interval model to calculate the optimal interval of the wastewater purification efficiency;

[0183] It should be noted that the role of calculating the optimal interval of the wastewater purification efficiency is as follows:

[0184] Function 1: Provide a dynamic benchmark for efficiency evaluation. By using the quantile regression algorithm to correlate the historical fluctuation feature vector with the purification efficiency, a dynamic range covering specific quantiles is constructed to provide an evaluation benchmark that fits the actual working conditions for real-time efficiency monitoring.

[0185] Function 2: Support the decision-making of flow rate control. The optimization range is directly linked to the flow rate control model: If the real-time efficiency deviates from the optimization range, it provides an adjustment target for the closed-loop control model to ensure that the fine-tuning or gradient adjustment of the flow rate is always oriented towards maintaining the efficiency within the range, avoiding system oscillations caused by blind adjustment.

[0186] Exemplarily, for example, quantiles are taken as 0.05 and 0.95 respectively, and calculate and , and the range is covering the fluctuation feature vectors and wastewater purification efficiencies of the historical stable fluidization state from 5% to 95%. The extreme values at both ends of the data can be excluded through the control range;

[0187] If there are additional constraints in the process, such as the lower limit of the wastewater purification efficiency being 85%, then the adjusted range is ;

[0188] Those skilled in the art can understand that by correlating the wastewater purification efficiency with the fluctuation features in the stable fluidization state through quantile regression, the interval boundaries are dynamically adjusted with the boundaries of the fluctuation features. For example, when FD increases, that is, the complexity of the pressure drop signal increases, the quantile model is automatically updated , reflecting the change trend of the wastewater purification efficiency and making the optimization range fit the actual working conditions of wastewater purification;

[0189] Obtain the wastewater purification efficiency corresponding to the fluctuation features of the real-time fluidization state, compare it with the wastewater purification efficiency in the optimization range, and judge whether the wastewater purification efficiency in the stable fluidization state is within the optimization range.

[0190] Step 4: If not, judge whether the deviation of the wastewater purification efficiency from the optimization range is caused by the flow rate. If so, construct a flow rate control model to control and adjust the flow rate;

[0191] The method for judging whether the deviation of the wastewater purification efficiency from the optimization range is caused by the flow rate is:

[0192] Those skilled in the art can understand that in the wastewater purification fluidized bed reactor, the flow rate is the key factor dominating the pressure drop in the reactor;

[0193] When fluid passes through the quartz sand carrier bed, the flow rate directly affects the frictional resistance and fluidization state of the solid-liquid two-phase: in the turbulent state, the pressure drop increases approximately quadratically with the increase of the flow rate (in line with fluid mechanics theories such as the Ergun equation), and it is necessary to maintain uniform fluidization of particles through the flow rate to ensure efficient mass transfer. Although the characteristics of the carrier particles (particle size, porosity), physical properties of the fluid (viscosity, density), and reactor structure (inverted cone hopper, packing layer) will affect the pressure drop, as the core operating variable, the dynamic regulation of the flow rate directly determines the magnitude and stability of the bed resistance;

[0194] Through the causal method of event sequence, combined with other interference factors of the wastewater treatment process corresponding to the optimization interval, excluding other interference factors of the current wastewater treatment process, misjudgment when the wastewater purification efficiency in the stable fluidization state caused by the flow rate is not in the optimization interval, that is, judging whether the wastewater purification efficiency deviates from the optimization interval due to the flow rate;

[0195] It should be explained that other interference factors of the wastewater treatment process include: initial COD, pollutant concentration, temperature, and pH, etc.;

[0196] When excluding the misjudgment of unqualified efficiency caused by the flow rate through the causal method of event sequence, it is necessary to first align the flow rate C, wastewater purification efficiency D, and interference factors (such as initial COD, pollutant concentration A / B) in time series to construct a multivariate time series model (such as vector autoregression VAR);

[0197] By using the Granger causality test in the causal method of event sequence to control the influence of A / B, verifying the independent prediction ability of C on D; then comparing the association between D and D under normal and abnormal states of A / B layer by layer, and using methods such as propensity score matching to quantify the causal effect of C on D, stripping the confounding influence of A / B; at the same time, mining high-frequency event chain patterns (such as the time series pattern of "C abnormal → D unqualified" accompanied by normal A / B), combined with counterfactual tests (conditional probability of assuming Y is qualified when X is normal), finally, through statistical significance and physical time series logic, determining whether the flow rate is an independent influencing factor, achieving effective exclusion of other interference factors, and ensuring the reduction of the misjudgment rate;

[0198] Through causal inference and hierarchical control of interference factors, the true influence of the flow rate on the efficiency can be located;

[0199] If the wastewater purification efficiency deviates from the optimization interval due to the flow rate, determine the optimization range of the flow rate through the optimization interval ;

[0200] Set dynamic buffer boundaries for the optimization range of the flow rate to reduce the impact of frequent adjustment of the flow rate on the fluidized bed reaction, and mark the optimization range with dynamic buffer boundaries as ;

[0201] Exemplarily, if the dynamic buffer boundary is 5%, the optimal flow rate range of the dynamic buffer boundary is set ;

[0202] Construct a flow rate control model, and input the optimal flow rate range and the dynamic buffer boundary into the flow rate control model to achieve the control and adjustment of the flow rate;

[0203] Among them, the construction method of the flow rate control model is as follows:

[0204] A1. Construct a closed-loop control model based on the real-time flow rate, wastewater purification efficiency, and interference factors (initial COD, pollutant concentration);

[0205] It should be explained that the model predictive controller is used to stabilize the flow rate within the optimal flow rate range while compensating for the minimum flow rate and reducing the impact of fluctuations in the initial COD and pollutant concentration on the wastewater purification efficiency;

[0206] A2. Establish a hierarchical response mechanism to achieve the control of the flow rate;

[0207] Preferably, the primary control: when the wastewater purification efficiency is close to the dynamic buffer boundary (such as 90% or 95%) of the optimal interval but does not exceed it, gradually pull the flow rate back to the optimal flow rate range by fine-tuning the frequency of the inlet / circulation pump (the adjustment amplitude each time ≤ 3%);

[0208] The secondary control: when the wastewater purification efficiency continuously deviates from the optimal interval for more than a preset time (such as 30 minutes), and causal analysis confirms that the flow rate is the main reason, start the flow rate gradient adjustment strategy (such as 5% - 10% each time), and record the efficiency response curves before and after the adjustment at the same time to avoid over-adjustment causing the oscillation of the fluidized bed reactor;

[0209] It should be explained that the primary control is used to fine-tune the flow rate, and the secondary control is used to quickly control the flow rate.

[0210] The technical solution of this embodiment is: based on the determined fluidization state, construct a fluctuation feature vector and identify the sub-states of the fluid state change. If the sub-state is a stable fluidization state, identify whether the wastewater purification efficiency in the stable fluidization state is within the optimal interval; if not, judge whether the wastewater purification efficiency deviates from the optimal interval due to the flow rate. If so, construct a flow rate control model to control and adjust the flow rate. After confirming that it is a flow rate problem, set the dynamic buffer boundary, adopt a hierarchical response mechanism combined with a closed-loop control model to achieve flow rate control, reduce the possibility of mass transfer failure or carrier loss caused by improper flow rate, and maintain the optimal degradation efficiency in the stable fluidization state.

[0211] Example Six

[0212] As Figure 3As shown in the figure, a liquid flow rate monitoring system further includes the following modules:

[0213] Curve construction module: Collect the fluid flow rate and pressure drop in S2 in real time, and establish a change curve of the real-time pressure drop and the rising flow rate;

[0214] Preferably, an electromagnetic flowmeter is installed in the fluidized bed reactor in S2 to collect the rising flow rate of the fluid in the main fluidized bed in real time;

[0215] Pressure sensors are installed at the inlet and outlet of the fluidized bed reactor respectively to measure the pressure drop between the inlet and outlet of the fluidized bed reactor;

[0216] Based on the rising flow rate of the fluid and the pressure drop of the packing layer, and perform smoothing processing on the pressure drop and flow rate, and construct a change curve of the real-time flow rate and pressure drop.

[0217] State recognition module: By collecting the flow rate and fluid change characteristics in the historical process data, construct a fluidization recognition model, and input the change curve of the real-time pressure drop and the rising flow rate into the fluidization recognition model to identify the fluidization state of the fluidized bed;

[0218] Obtain the rising flow rate and pressure drop of the photo-Fenton coupled fluidized bed reaction from the historical process data, perform smoothing processing on the pressure drop and flow rate in the historical data, identify and process the outliers in the historical data through the 3σ criterion, and construct a change curve of the historical flow rate and pressure drop;

[0219] Verify the change curve of the historical flow rate and pressure drop by Darcy's law, and identify the change curve of the historical flow rate and pressure drop as the laminar flow section and the turbulent flow section through the flow rate demarcation point;

[0220] Among them, the method of verifying Darcy's law is:

[0221] Identify the inflection point of the change curve of the historical flow rate and pressure drop, divide the change curve of the historical flow rate and pressure drop into two sections based on the inflection point, and obtain the low flow rate section and the high flow rate section;

[0222] Perform fitting verification of Darcy's law on the low flow rate section, and the fitting model is: ;

[0223] Perform polynomial regression fitting verification on the high flow rate section, and the fitting model is: ;

[0224] Among them, is the pressure drop, is the fluid flow rate, and a, b, c, d, and e are the coefficients of the fitting model respectively;

[0225] If there are multiple inflection points on the change curve of historical flow rate and pressure drop, obtain the root mean square error of fitting verification in the low flow rate stage and the root mean square error of verification in the high flow rate stage corresponding to each inflection point;

[0226] Construct a stage error pair from the root mean square error of fitting verification in the low flow rate stage and the root mean square error of fitting verification in the high flow rate stage corresponding to each inflection point;

[0227] Obtain the stage error pairs of all inflection points, and select the inflection point with the smallest root mean square error in the stage error pairs as the flow rate demarcation point;

[0228] Based on the flow rate demarcation point, divide the change curve of historical flow rate and pressure drop into a laminar flow section and a turbulent flow section;

[0229] Based on the change curves of historical flow rate and pressure drop in the laminar flow section and the turbulent flow section, construct a fluidization identification model, and use the Ergun equation to verify and correct the fluidization identification model;

[0230] Among them, the Ergun equation is: , L is the bed height, dl is the particle diameter in the fluidized bed, n is the fluid viscosity, is the fluid density, is the fluid flow rate, is the bed porosity;

[0231] Obtain the change curve of real-time pressure drop and rising flow rate, input it into the fluidization identification model, and identify the fluidization state of the fluidized bed.

[0232] Optimization discrimination module: Based on the determined fluidization state, construct a fluctuation feature vector and identify the sub-states of fluid state change. If the sub-state is a stable fluidization state, identify whether the wastewater purification efficiency in the stable fluidization state is in the optimization interval;

[0233] Based on the determined fluidization state, construct a fluctuation feature vector and identify the sub-states of fluid state change;

[0234] Preferably, the sub-states include a stable fluidization state, a chaotic fluidization state, and a critical fluidization state;

[0235] Among them, the construction method of the fluctuation feature vector is:

[0236] Calculate the peak factor CF, power spectrum entropy PSE, fractal dimension FD, and Lyapunov exponent LE of the pressure drop within the monitoring period;

[0237] Preferably, the acquisition method of the peak factor CF is:

[0238] Through the formula: Obtain the peak factor CF, where is the pressure drop collected in real time by the pressure sensor within the monitoring period, The maximum pressure drop and the minimum pressure drop within the monitoring period are respectively the average pressure drop within the monitoring period;

[0239] Preferably, the power spectrum entropy PSE is obtained by:

[0240] Through the formula: Obtain the power spectrum entropy PSE;

[0241] Among them, the way to obtain PS i is:

[0242] Perform a fast Fourier transform on the pressure drop in the monitoring period, convert the pressure drop in the time domain into the pressure drop in the frequency domain, and obtain the discrete frequency component f i corresponding power S(f i );

[0243] Calculate the proportion of the power of each discrete frequency component in the total power to obtain PS i , where i is the number of the frequency component and n is the total number of frequency components;

[0244] Use the box dimension method to calculate the fractal dimension of the pressure drop signal to obtain the fractal dimension FD;

[0245] Calculate the maximum Lyapunov exponent of the pressure drop signal as the Lyapunov exponent LE;

[0246] Based on the peak factor CF, power spectrum entropy PSE, fractal dimension FD, and Lyapunov exponent LE as fluctuation characteristics, construct a fluctuation characteristic vector ;

[0247] Based on the random forest algorithm, input the fluctuation characteristic vector into the random forest algorithm to realize the identification of the sub-states of the fluidization state;

[0248] If the sub-state is the stable fluidization state, the method for identifying whether the wastewater purification efficiency in the stable fluidization state is in the optimal range is:

[0249] Obtain the wastewater purification efficiency characteristics in the stable fluidization state in the historical data, including the COD removal rate and the pollutant degradation rate;

[0250] At the same time, extract the fluctuation characteristic vector in the historical data, and construct a dynamic interval model for the fluctuation characteristic vector in the historical data and the wastewater purification efficiency characteristics in the stable fluidization state in the historical data through the quantile regression algorithm;

[0251] Through the formula: , establish a quantile model , where X is the fluctuation characteristic vector in the historical data, is the wastewater purification efficiency characteristic, is the quantile, the corresponding quantile of the regression coefficient vector, which is obtained by performing quantile regression training on historical data;

[0252] Input the fluctuation feature vector of the historical stable fluidization state and the wastewater purification efficiency into the dynamic interval model, and calculate the optimization interval of the wastewater purification efficiency;

[0253] Obtain the wastewater purification efficiency corresponding to the real-time fluctuation characteristics of the fluidization state, compare it with the wastewater purification efficiency in the optimization interval, and judge whether the wastewater purification efficiency in the stable fluidization state is within the optimization interval.

[0254] Control adjustment module: If not, judge whether the deviation of the wastewater purification efficiency from the optimization interval is caused by the flow rate. If so, construct a flow rate control model to control and adjust the flow rate;

[0255] The method for judging whether the deviation of the wastewater purification efficiency from the optimization interval is caused by the flow rate is as follows:

[0256] Through the event sequence causality method, combined with other interference factors of the wastewater treatment process corresponding to the optimization interval, exclude other interference factors of the current wastewater treatment process, and misjudge when the flow rate causes the wastewater purification efficiency in the stable fluidization state not to be within the optimization interval, that is, judge whether the deviation of the wastewater purification efficiency from the optimization interval is caused by the flow rate;

[0257] If the deviation of the wastewater purification efficiency from the optimization interval is caused by the flow rate, determine the optimization range of the flow rate through the optimization interval;

[0258] Set dynamic buffer boundaries for the optimization range of the flow rate to reduce the impact of frequent flow rate adjustment on the fluidized bed reaction;

[0259] Construct a flow rate control model, input the optimization range of the flow rate and the dynamic buffer boundaries into the flow rate control model, and realize the control and adjustment of the flow rate.

[0260] Example Seven

[0261] As Figure 4 shown, the photo-Fenton coupled fluidized bed reaction device includes the following modules:

[0262] Photocatalytic reaction module: Using an ultraviolet mercury lamp as the light source, arranged on the inner wall or central axis of the reactor, with a filler of photocatalyst, generating electron-hole pairs by light energy excitation of the photocatalyst, promoting the generation of hydroxyl radicals and driving the Fenton cycle;

[0263] Fluidized bed reactor module: It includes an inverted cone hopper, a packing reaction zone and a sedimentation separation zone. The liquid flow rate is maintained by a circulation pump to make the packing loaded with catalyst in a fluidized state, realizing solid-liquid mixing and mass transfer. The design of the inverted cone hopper and the sedimentation zone promotes solid-liquid separation;

[0264] Fenton reagent dosing module: It includes a hydrogen peroxide and ferrous ion dosing unit and a pH adjustment unit to create an acidic condition for the Fenton reaction;

[0265] Three-phase separation module: Located at the top of the reactor, it separates the treated water, reaction gas and carrier particles through inclined plates and a collecting trough, realizing efficient three-phase separation of solid, liquid and gas;

[0266] Circulation system module: It consists of a circulation pump, a reflux pipeline and a flow meter to control the liquid reflux ratio and promote uniform mixing of the reagent and the wastewater;

[0267] Auxiliary equipment module: It includes an adjustment tank, a sedimentation tank and a sludge treatment unit to assist the main device in stable operation and treat by-products.

[0268] The above has described a detailed embodiment of the present invention, but the content described is only a preferred embodiment of the present invention and cannot be considered as used to limit the scope of implementation of the present invention. All equivalent changes and improvements made according to the scope of the application of the present invention should still fall within the scope covered by the patent of the present invention.

Claims

1. Photocatalytic-Fenton coupled fluidized bed wastewater treatment process, characterized in that, It includes the following steps: Wastewater regulation and preliminary chemical dosing; Photocatalysis-fluidized bed coupling reaction and flow rate monitoring; Solid-liquid-gas three-phase separation and catalyst recovery; Advanced treatment and sludge disposal; The method of flow rate monitoring is as follows: Collect the fluid flow rate and pressure drop, and establish the change curve of the real-time pressure drop and the rising flow rate; By collecting the flow rate and fluid change characteristics in the historical process data, construct a fluidization identification model, and input the change curve of the real-time pressure drop and the rising flow rate into the fluidization identification model to identify the fluidization state of the fluidized bed; Based on the determined fluidization state, construct a fluctuation feature vector and identify the sub-states of the fluid state change. If the sub-state is a stable fluidization state, identify whether the wastewater purification efficiency in the stable fluidization state is within the optimal range; If it is not within the range, judge whether the wastewater purification efficiency deviates from the optimal range due to the flow rate. If so, construct a flow rate control model to control and adjust the flow rate.

2. The photo-Fenton coupled fluidized bed wastewater treatment process according to claim 1, characterized in that, The method of constructing the fluidization identification model is as follows: Smooth the pressure drop and flow rate in the historical data, identify and process the outliers in the historical data, and construct the change curve of the historical flow rate and pressure drop; Verify the change curve of the historical flow rate and pressure drop by Darcy's law, identify the change curve of the historical flow rate and pressure drop through the flow rate demarcation point, and divide the change curve into a laminar flow section and a turbulent flow section; Based on the change curves of the historical flow rate and pressure drop in the laminar flow section and the turbulent flow section, construct a fluidization identification model, and use the Ergun equation to verify and correct the fluidization identification model.

3. The photo-Fenton coupled fluidized bed wastewater treatment process according to claim 2, characterized in that, The method of obtaining the flow rate demarcation point is as follows: Identify the inflection point of the change curve of the historical flow rate and pressure drop, and divide the change curve of the historical flow rate and pressure drop into two sections based on the inflection point to obtain a low flow rate section and a high flow rate section; Conduct fitting verification on the low flow rate section by Darcy's law, and conduct polynomial regression fitting verification on the high flow rate section; Analyze the fitting errors of the low flow rate section and the high flow rate section, construct a stage error pair, and select the inflection point with the minimum root mean square error in the stage error pair as the flow rate demarcation point.

4. The photo-Fenton coupled fluidized bed wastewater treatment process according to claim 3, characterized in that, The method of analyzing the fitting errors of the low flow rate section and the high flow rate section is as follows: If there are multiple inflection points in the change curve of the historical flow rate and pressure drop, obtain the root mean square error of the fitting verification in the low flow rate stage and the root mean square error of the fitting verification in the high flow rate stage corresponding to each inflection point; Construct a stage error pair with the root mean square error of the fitting verification in the low flow rate stage and the root mean square error of the high flow rate stage verification corresponding to each inflection point.

5. The photo-Fenton coupled fluidized bed wastewater treatment process according to claim 1, characterized in that, The method of identifying the sub-states of the fluid state change is as follows: Based on the determined fluidization state, construct a fluctuation feature vector and identify the sub-states of the fluid state change; Based on the random forest algorithm, input the fluctuation feature vector into the random forest algorithm to realize the identification of the sub-states of the fluidization state.

6. The photo-Fenton coupled fluidized bed wastewater treatment process according to claim 5, characterized in that, The method of constructing the fluctuation feature vector is as follows: Calculate the peak factor, power spectrum entropy, fractal dimension, and Lyapunov exponent of the pressure drop within the monitoring period, and construct a fluctuation feature vector.

7. The photo-Fenton coupled fluidized bed wastewater treatment process according to claim 1, characterized in that The method of identifying whether the wastewater purification efficiency in the stable fluidization state is within the optimal range is as follows: Obtain the wastewater purification efficiency characteristics in the stable fluidization state in the historical data, and at the same time extract the fluctuation feature vector in the historical data; Construct a dynamic interval model by using the quantile regression algorithm for the fluctuation feature vector in historical data and the wastewater purification efficiency feature in the stable fluidization state in historical data; Input the fluctuation feature vector and wastewater purification efficiency in the historical stable fluidization state into the dynamic interval model, and calculate the optimization interval of the wastewater purification efficiency; Obtain the wastewater purification efficiency corresponding to the fluctuation feature of the real-time fluidization state, compare it with the wastewater purification efficiency in the optimization interval, and determine whether the wastewater purification efficiency in the stable fluidization state is within the optimization interval.

8. The photo-Fenton coupled fluidized bed wastewater treatment process according to claim 1, characterized in that The method for controlling and adjusting the flow rate is as follows: If the wastewater purification efficiency deviates from the optimization interval due to the flow rate, determine the optimization range of the flow rate through the optimization interval, and set a dynamic buffer boundary for the optimization range of the flow rate; Construct a flow rate control model, input the optimization range of the flow rate and the dynamic buffer boundary into the flow rate control model, and realize the control and adjustment of the flow rate.

9. The photo-Fenton coupled fluidized bed wastewater treatment process according to claim 8, wherein, The determination method for "if the wastewater purification efficiency deviates from the optimization interval due to the flow rate" is as follows: Through the event sequence causality method, combined with other interference factors of the wastewater treatment process corresponding to the optimization interval, determine whether the wastewater purification efficiency deviates from the optimization interval due to the flow rate.

Images