Enhanced coagulation dosing method and system based on waterworks micro-plastic treatment

Abstract

The invention relates to the technical field of micro-plastic treatment, in particular to an enhanced coagulation dosing method and system based on tap water plant micro-plastic treatment, and the method comprises the following steps: collecting a first water quality parameter of raw water at a water inlet of a tap water plant; inputting the first water quality parameter into a BP neural network feedforward controller, and predicting the adding amount of a flocculating agent; a micro-plastic abundance value of a water outlet and a corresponding abundance set value are used as input linguistic variables, and a fuzzy PID feedback controller is used for controlling the adding amount of a flocculating agent; and detecting the turbidity of the outlet of the sedimentation tank, inputting the second water quality parameter into the post-dosing BP neural network controller when the turbidity value of the water outlet does not meet the set condition, and outputting the adding amount of the flocculating agent in the water outlet pipe of the sedimentation tank. By adopting an intelligent feedforward-feedback composite dosing and secondary dosing control strategy, quick response and accurate dosing can be realized, the medicine consumption and the labor cost are reduced, and micro-plastics and turbidity in tap water are effectively controlled.

Description

Technical Field

[0001] The present invention relates to the technical field of microplastic treatment, and in particular to an enhanced coagulation and dosing method and system based on microplastic treatment in a water plant. Background Art

[0002] Plastic products are widely used in human production and life, and discarded plastics enter water bodies in large quantities in the form of plastic particles or fragments. Microplastics are small in size, large in specific surface area, and difficult to degrade. They easily absorb persistent organic pollutants, heavy metals and pathogenic microorganisms, and are considered to be new pollutants that pose high risks to the ecological environment and human health.

[0003] At present, the removal of microplastics mainly adopts technologies such as coagulation, filtration, adsorption, advanced oxidation and membrane separation; the conventional treatment process of water plants is filtration and disinfection after coagulation and sedimentation, which mainly removes suspended matter, colloids and dissolved organic matter, and there is no special treatment process for microplastics. Studies have shown that coagulation and filtration also have a removal effect on microplastic particles, but the removal rate varies greatly due to the influence of raw water quality, process conditions and the properties of microplastics, and there is a lack of effective control of microplastics. Especially in the case of manual operation, the addition of flocculants is relatively rough, and in order to ensure water quality, excessive flocculants are often added, resulting in increased drug consumption and increased aluminum ion concentration in the effluent. Therefore, process intelligent control has become a necessary way to improve production efficiency, reduce energy consumption and reduce labor costs.

[0004] The publication number is CN118458907A. A DNN model is constructed, and the input layer is constructed according to the water inlet conditions and real-time water quality characteristics. However, the feature support of the DNN model category is poor: DNN does not support category features well. As the number of category features increases, dimensionality explosion may occur.

[0005] The publication number is CN118331209A. A radial basis function neural network is used for pre-training to obtain a mapping relationship model. The sewage quality data, sewage volume, sewage quality data at the factory and the type of reagents are used as input data of the radial basis function neural network, and the dosage prediction value is used as the output data of the radial basis function neural network to construct a mapping relationship model. However, the radial basis function neural network has problems with the number of hidden layer neurons, parameter selection, training process and local optimality. Summary of the invention

[0006] In view of the shortcomings of the existing methods, the present invention first uses a BP neural network feedforward controller to predict the amount of flocculant, and then uses a fuzzy PID feedback controller to adjust the flow of the metering pump according to the microplastic abundance value, thereby adjusting the dosage of flocculants and coagulants in the flocculation reaction tank and sedimentation tank.

[0007] The technical solution adopted by the present invention is: a method for enhancing coagulation and dosing based on microplastic treatment in a water plant comprises the following steps:

[0008] Step 1, collecting the first water quality parameter of raw water at the water inlet of the water plant;

[0009] As a preferred embodiment of the present invention, the first water quality parameter includes: raw water flow, turbidity, microplastic abundance, pH value and water temperature.

[0010] Step 2: input the first water quality parameter into the BP neural network feedforward controller to predict the flocculant dosage;

[0011] As a preferred implementation of the present invention, the BP neural network feedforward controller is a three-layer BP neural network model.

[0012] Step 3: Using the abundance value of microplastics at the outlet of the sedimentation tank and the corresponding abundance setting value as input language variables, the dosage of the flocculant is controlled by a fuzzy PID feedback controller;

[0013] As a preferred embodiment of the present invention, step three includes:

[0014] The deviation e between the set value and the measured value of microplastic abundance before filtration and the deviation change rate de / dt are used as input linguistic variables of the fuzzy PID feedback controller; the proportional coefficient Kp, the integral coefficient Ki and the differential coefficient Kd are used as output linguistic variables; the variable domain set is fuzzy processed through the membership function, and the self-tuning rules are summarized in combination with actual engineering experience to correct the PID parameters in real time, thereby correcting the flocculant dosage.

[0015] As a preferred implementation of the present invention, step three also includes: determining the flocculant dosage in the pipeline mixer by integrating the calculation results of the BP neural network controller and the fuzzy PID controller.

[0016] As a preferred embodiment of the present invention, it also includes: determining the dosage of the coagulant aid according to the dosage of the flocculant.

[0017] Step 4: Detect the turbidity of the sedimentation tank outlet. When the turbidity value of the outlet does not meet the set conditions, input the second water quality parameter into the post-dosing BP neural network controller and output the flocculation dosage to the sedimentation tank outlet pipe.

[0018] As a preferred embodiment of the present invention, the second water quality parameter includes: water flow rate, turbidity, microplastic abundance, pH value and water temperature.

[0019] As a preferred embodiment of the present invention, the post-drug addition BP neural network controller comprises:

[0020] The turbidity setting value of the sedimentation tank outlet is Z, and the measured value is Z * , if Z *≤Z, no operation is performed; if Z*>Z, the post-dosing BP neural network controller takes the second water quality parameter as the input parameter, turbidity as the control amount, and the flocculant dosage as the output layer node.

[0021] As a preferred embodiment of the present invention, a system for an enhanced coagulation and dosing method for microplastic treatment in a water plant includes:

[0022] BP neural network feedforward controller and fuzzy PID feedback controller, post-dosing BP neural network controller and actuator;

[0023] Among them, the BP neural network feedforward controller, the fuzzy PID feedback controller and the post-dosing BP neural network controller are set in the PLC controller with programming function; the PLC controller collects and processes the monitoring data, performs logical operations and numerical operations, and transmits the processed signals to the actuator; and the flow rate of the metering pump is controlled by changing the frequency of the variable frequency actuator.

[0024] Beneficial effects of the present invention:

[0025] 1. The composite control of BP neural network feedforward controller and fuzzy PID feedback controller has the advantages of high control accuracy and fast stabilization speed. It can respond quickly when the raw water changes, ensuring the accuracy of dosing and stable effluent water quality;

[0026] 2. Use the post-dosing BP neural network controller to add flocculants again to the outlet pipe of the sedimentation tank to control the turbidity of the outlet, reduce the burden on the filter tank, and ensure the quality of the effluent;

[0027] 3. The turbine dynamic pipeline mixer can enhance the mixing effect of the reagents, improve the reaction efficiency of the reagents, and reduce the dosage of the reagents;

[0028] 4. The micro-vortex flocculation process creates highly efficient coagulation and flocculation hydraulic conditions, and has the advantages of short flocculation time, good flocculation effect, good floc sedimentation performance, and low drug consumption; the process is less affected by shock loads, has stable operation, high water production rate, and has application and promotion value;

[0029] 5. Through the compounding of flocculants and coagulant aids, the bridging, electrical neutralization and adsorption synergistic effects of organic-inorganic flocculants are brought into play, the coagulation effect is enhanced, the dosage is reduced and the cost is reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a flow chart of the enhanced coagulation and dosing method based on microplastic treatment in a water plant of the present invention;

[0031] Figure 2 It is the block diagram of BP neural network feedforward-fuzzy PID feedback composite system;

[0032] Figure 3 It is a schematic diagram of pipeline mixing and flocculation sedimentation tank. DETAILED DESCRIPTION

[0033] The present invention is further described below in conjunction with the accompanying drawings and embodiments. This figure is a simplified schematic diagram, which only illustrates the basic structure of the present invention in a schematic manner, and therefore it only shows the components related to the present invention.

[0034] like Figure 1 As shown, a method for enhanced coagulation and dosing based on microplastic treatment in a water plant includes the following steps:

[0035] The coagulation process is a key link in the water production process. The coagulation process is operated using automated control, including a water quality monitoring unit, a feedforward-feedback composite control unit and a dosing execution unit. Data transmission is achieved between the units through network connections.

[0036] Step 1, collecting the first water quality parameter of raw water at the water inlet of the water plant;

[0037] The first water quality parameters include: raw water flow, turbidity, microplastic abundance, pH value and water temperature;

[0038] Step 2: input the first water quality parameter into the BP neural network feedforward controller to predict the flocculant dosage;

[0039] The BP neural network feedforward controller and the fuzzy PID feedback controller are combined to form a feedforward-feedback composite control system;

[0040] The BP neural network feedforward controller adopts a three-layer BP neural network model, takes water quality parameters as the input layer neurons of the BP neural network, the abundance of microplastics in the sedimentation tank effluent as the control quantity, and the flocculant dosage as the output quantity; the historical operation data of the water plant is divided into training data and test data to complete the learning function of the BP neural network feedforward controller; the trained BP neural network feedforward controller inputs the on-site water quality parameters into the controller to calculate the preliminary flocculant dosage.

[0041] Step 3: Using the microplastic abundance value at the outlet and the corresponding abundance setting value as input language variables, a fuzzy PID feedback controller is used to control the dosage of the flocculant;

[0042] The fuzzy PID feedback controller uses the deviation e and the deviation change rate de / dt between the set value and the measured value of the microplastic abundance before filtration as the input linguistic variables of the fuzzy PID feedback controller; the proportional coefficient Kp, the integral coefficient Ki and the differential coefficient Kd are used as the output linguistic variables; the variable domain set is fuzzy processed by the membership function, and the membership functions include: triangular, trapezoidal and Gaussian membership functions; combined with actual engineering experience, the self-tuning rules of the controlled process for the proportional coefficient change ΔKp, the integral coefficient change ΔKi and the differential coefficient change ΔKd under different working conditions are summarized, and the fuzzy rule table of the three key parameters is obtained; during actual operation, the fuzzy PID feedback controller looks up the table according to the current actual e and de / dt, obtains the ΔKp, ΔKi and ΔKd corresponding to the real-time control, and corrects the PID parameters to achieve the optimal state, thereby adjusting the flocculant dosage.

[0043] The PLC controller integrates the calculation results of the BP neural network feedforward controller and the fuzzy PID feedback controller, transmits the signal to the frequency converter 1 at one end of the pipeline mixer, and controls the metering pump 1 to add flocculant to the pipeline mixer through the frequency converter 1; the raw water and flocculant are quickly mixed in the pipeline mixer and then enter the flocculation reaction tank.

[0044] A coagulant dosing port is installed at the rear 1 / 3 section of the flocculation reaction tank, corresponding to frequency converter 2 and metering pump 2; the coagulant dosage is determined according to the flocculant dosage data transmitted by the feedforward-feedback composite control system; the coagulant dosage is set to a certain proportion of the flocculant dosage, and the proportion can be customized by each water plant; the PLC controller transmits the signal to frequency converter 2, thereby controlling metering pump 2 to achieve accurate dosing of the coagulant.

[0045] Step 4: Detect the turbidity of the sedimentation tank outlet. When the turbidity value of the outlet does not meet the set conditions, input the second water quality parameter into the post-dosing BP neural network controller and output the flocculation dosage to the sedimentation tank outlet pipe.

[0046] The turbidity setting value of the sedimentation tank outlet is Z, and the measured value is Z * , if Z *≤Z, no operation is performed; if Z*>Z, the post-dosing BP neural network controller is started; the post-dosing BP neural network controller is a BP neural network controller set at the outlet of the sedimentation tank, and the outlet water flow, turbidity before filtration, pH value, and water temperature signal of the sedimentation tank outlet are used as input parameters, turbidity is the control quantity, and the flocculant dosage is the output layer node. The learning function of the neural network controller is completed with the historical operation data of the water plant. The trained neural network calculates the dosage of the post-dosing flocculant through the input of parameters such as the outlet water flow, turbidity, pH value, and water temperature; the PLC controller transmits the dosage signal to the frequency converter 3, thereby controlling the metering pump 3 to achieve accurate dosing of the flocculant. The flocculant addition point is set at the outlet of the sedimentation tank, and is directly micro-flocculated and then filtered.

[0047] The control principle is: train the BP neural network through the big data of water plant operation to enable it to have adaptive learning ability, and be able to continuously and accurately predict the dosage of flocculant solution according to the changes in raw water quality, thereby improving the control accuracy of the automatic dosing system; at the same time, use the fuzzy PID algorithm to dynamically optimize the PID controller parameters according to the sedimentation tank effluent, automatically adjust the dosage of flocculant, ensure optimal control of the dosage of the agent, and solve the problem of system oscillation out of control caused by drastic changes in raw water quality and quantity, or untimely dosing resulting in effluent that does not meet the requirements.

[0048] like Figure 2 and 3 As shown, the system based on the enhanced coagulation and dosing method for microplastic treatment in water plants includes:

[0049] The monitoring unit is used to obtain real-time water quality and quantity parameters. The monitoring instruments include flow meters, temperature sensors, online pH meters, turbidity sensors, flow cytometers, etc.; the flow rate, water temperature, pH value, turbidity and microplastic abundance are detected respectively; the microplastic water samples are concentrated and filtered and then fluorescently labeled, and the microplastic particles are quickly identified and counted with the help of high-throughput flow cytometer automatic image recognition technology; monitoring points are set at the raw water inlet, sedimentation tank outlet and carbon filter tank outlet, and the instrument takes samples at fixed intervals.

[0050] The feedforward-feedback composite control unit is composed of a BP neural network feedforward controller and a fuzzy PID feedback controller; the control equipment uses a PLC system to complete the collection and processing of on-site monitoring data, perform logical operations and numerical operations, and transmit the processed signals to the actuator to control its working state; that is, the calculated values ​​of flow value and frequency are converted into electrical signals to control the opening and closing of the valve and the frequency of the inverter, thereby controlling the flow of the metering pump. Among them, the actuator includes: inverter and metering pump.

[0051] The present invention follows the traditional water treatment process of coagulation sedimentation plus filtration and disinfection, and achieves the goals of precise drug dosing, stable water supply and energy saving and consumption reduction by strengthening the coagulation and deep filtration processes and combining them with an intelligent drug dosing system.

[0052] The water production process is: dynamic pipeline mixer, micro-vortex flocculation tank, inclined plate sedimentation tank, sand filter, carbon filter and disinfection tank.

[0053] In order to enhance the mixing of reagents, a turbine dynamic pipeline mixer is used instead of a conventional mixing tank, and a vortex pipeline mixer is installed within 20-50 meters from the front end of the flocculation reaction tank; a turbine is installed inside the pipeline, and the rotation of the turbine in the pipeline is controlled by an external motor; the turbine speed can be automatically adjusted according to the inlet flow rate and temperature to ensure that the water flow velocity gradient value G required for mixing under different working conditions is achieved; the rotation of the turbine causes the water flow in the pipeline to spiral forward to form a vortex of a certain distance, so that the reagent and the water body are quickly and evenly mixed, thereby improving the reaction efficiency of the reagent, reducing the dosage of the reagent, accelerating the formation of alum flocs and optimizing the floc sedimentation effect.

[0054] The flocculant used is polyaluminium ferric chloride (PAFC) or polyaluminium chloride, and the coagulant aid is polyacrylamide (PAM); several dosing points are set within 1-5 meters before the pipeline mixer, the front dosing point of the flocculant is added from the pipeline mixer dosing point, and the rear dosing point is added from the outlet of the inclined plate sedimentation tank; the coagulant aid is added from multiple points in the rear section of the flocculation reaction tank; if necessary, the pre-oxidant or other additives can also be added from the pipeline mixer dosing point.

[0055] In order to improve the efficiency of coagulation reaction, the micro-vortex flocculation process is adopted, and its core is to fill the micro-vortex reactor in the flocculation reaction tank; the micro-vortex reactor is a porous hollow sphere structure, and the porosity and pore size are determined according to the process requirements, and the inner and outer surfaces are roughened; the vortex reactor has no directionality and can be directly put into water for accumulation without blocking the wall holes, and does not require fixed installation; the water flow repeatedly passes through the small holes of the micro-vortex reactor, and the friction resistance with the inner and outer walls changes its through-hole flow velocity and flow direction, causing the water flow to generate a large number of tiny vortex flows; the tiny vortex flow can effectively promote the diffusion and collision of particles in the water, and has a higher coagulation efficiency; in addition, a floc suspension zone is formed inside the micro-vortex reactor, which forms a three-dimensional contact flocculation of the destabilized colloid, generating high-density alum flocs with good sedimentation performance.

[0056] In order to prevent the loss of the vortex reactor, a partition wall is set in the flocculation reaction tank to appropriately divide the tank into sections for fixing the micro-vortex reactor bracket. The bottom is supported by the bracket and the top is installed by a press net. The number of micro-vortex reaction balls filled in the flocculation reaction tank decreases in the direction of water flow until there is no ball filled in the last 1 / 3 of the reaction tank.

[0057] The flocculation tank is set up as follows: vortex reactors with different surface opening diameters and opening ratios are selected to control the flow rate of water through the holes. The flow rate through the holes of the front zone vortex reactor should be slightly greater than that of the rear zone vortex reactor, and the diameter of the front zone vortex reactor should be slightly smaller than that of the rear zone vortex reactor.

[0058] The flocculation tank is set up as follows: the flocculation reaction tank and the sedimentation tank are built together, and the sedimentation tank adopts an inclined plate or inclined tube sedimentation tank; since the vortex reactor contains suspended sludge, sludge deposition occurs; a sludge discharge area needs to be set at the bottom of the combined tank to discharge the sludge outside the vortex reactor.

[0059] Both sand filters and carbon filters are deep filters with a filter layer thickness of 2.0 meters. The filter media for sand filters are quartz sand, manganese sand or magnetite sand; the filter media for carbon filters are anthracite, porous activated carbon, and biomass charcoal. When there are no conditions to build sand filters and carbon filters separately, carbon sand filters can be used instead.

[0060] Engineering application practice shows that compared with conventional mixing tanks, turbine dynamic pipeline mixers can save 15-35% of flocculant dosage and can adapt to low-temperature, low-turbidity influent and overload operation. The coagulation efficiency of the micro-vortex coagulation process is better than that of the traditional coagulation process, and the reaction time can be shortened to 5-8 minutes. The turbidity of the effluent from the sedimentation tank can be stabilized below 3NTU, and the turbidity after filtration is less than 1NTU, and the water output can be increased by 1-2 times. Actual operation cases of relevant water treatment plants at home and abroad show that the overall removal rate of microplastic particles can reach more than 80% through coagulation, sedimentation, filtration and other processes. The enhanced coagulation and filtration process plus the intelligent dosing method can obtain better effluent water quality than the ordinary coagulation and filtration process.

[0061] Based on the ideal embodiment of the present invention, relevant staff can make various changes and modifications without departing from the technical concept of the present invention. The technical scope of the present invention is not limited to the content in the specification, and its technical scope must be determined according to the scope of the claims.

Claims

1. A method for enhanced coagulation and dosing based on microplastic treatment in a water plant, characterized in that: The following steps are involved: Step 1, collecting the first water quality parameter of raw water at the water inlet of the water plant; Step 2: input the first water quality parameter into the BP neural network feedforward controller to predict the flocculant dosage; Step 3: Using the abundance value of microplastics at the outlet of the sedimentation tank and the corresponding abundance setting value as input language variables, the dosage of the flocculant is controlled by a fuzzy PID feedback controller; Step 4: Detect the turbidity of the sedimentation tank outlet. When the turbidity value of the outlet does not meet the set conditions, input the second water quality parameter into the post-dosing BP neural network controller to output the flocculant dosage of the sedimentation tank outlet pipe.

2. The enhanced coagulation and dosing method based on microplastic treatment in a water plant according to claim 1 is characterized in that: The BP neural network feedforward controller is a three-layer BP neural network model.

3. The enhanced coagulation and dosing method based on microplastic treatment in a water plant according to claim 1 is characterized in that: Step three includes: The deviation e between the set value and the measured value of microplastic abundance before filtration and the deviation change rate de / dt are used as input linguistic variables of the fuzzy PID feedback controller; the proportional coefficient Kp, the integral coefficient Ki and the differential coefficient Kd are used as output linguistic variables; the variable domain set is fuzzy processed through the membership function, and the self-tuning rules are summarized in combination with actual engineering experience to correct the PID parameters in real time, thereby correcting the flocculant dosage.

4. The enhanced coagulation and dosing method based on microplastic treatment in a water plant according to claim 3 is characterized in that: Step three also includes: determining the amount of flocculant added in the pipeline mixer by integrating the calculation results of the BP neural network controller and the fuzzy PID controller.

5. The enhanced coagulation and dosing method based on microplastic treatment in a water plant according to claim 4 is characterized in that: Determine the dosage of coagulant aid according to the flocculant dosage.

6. The enhanced coagulation and dosing method based on microplastic treatment in a water plant according to claim 1, characterized in that: The first water quality parameters include: raw water flow, turbidity, microplastic abundance, pH value and water temperature.

7. The enhanced coagulation and dosing method based on microplastic treatment in a water plant according to claim 1, characterized in that: Secondary water quality parameters included: effluent flow, turbidity, microplastic abundance, pH value and water temperature.

8. The enhanced coagulation and dosing method based on microplastic treatment in a water plant according to claim 1, characterized in that: Post-dosing BP neural network controller includes: The turbidity setting value of the sedimentation tank outlet is Z, and the measured value is Z * , if Z * ≤Z, no operation is performed; if Z*>Z, the post-dosing BP neural network controller takes the second water quality parameter as the input parameter, turbidity as the control amount, and the flocculant dosage as the output layer node.

9. A system for the enhanced coagulation and dosing method based on microplastic treatment in a water plant according to any one of claims 1 to 8, characterized in that: include: BP neural network feedforward controller and fuzzy PID feedback controller, post-dosing BP neural network controller and actuator; Among them, the BP neural network feedforward controller, the fuzzy PID feedback controller and the post-dosing BP neural network controller are set in the PLC controller with programming function; The PLC controller collects and processes monitoring data, performs logical operations and numerical operations, and transmits the processed signals to the actuator; the flow rate of the metering pump is controlled by changing the frequency of the variable frequency actuator.

Images