A method and system for dosing a phosphorus removal agent for chemical phosphorus removal in a sewage treatment plant

Abstract

The application relates to the technical field of sewage treatment, in particular to a phosphorus removal reagent adding method and system for chemical phosphorus removal in a sewage treatment plant. The method comprises the following steps: collecting operation parameters of the sewage treatment plant in real time; inputting the collected operation parameters into a pre-constructed dynamic physical and chemical model, simulating the dynamic change of a phosphorus removal process, and rollingly predicting a variation trajectory of effluent phosphorus concentration in a future set prediction time domain; the dynamic physical and chemical model comprises a precipitation module, an aging module and an adsorption module; a model predictive control algorithm is adopted to take the minimum phosphorus removal reagent adding amount and the stable and up-to-standard effluent phosphorus concentration in the future set prediction time domain as an optimization target, so as to obtain an optimal phosphorus removal reagent adding amount sequence in the future set prediction time domain. The application can reduce reagent consumption, sludge production and operation cost.

Description

Technical Field

[0001] This application relates to the field of wastewater treatment technology, and in particular to a method and system for adding phosphorus removal agents for chemical phosphorus removal in wastewater treatment plants. Background Technology

[0002] Chemical phosphorus removal is a key process for ensuring that the total phosphorus content in the effluent of wastewater treatment plants meets standards, and its automatic control is crucial for improving efficiency and reducing chemical consumption. Currently, common automatic control methods mainly fall into two categories: First, there are control methods based on chemometrics and feedback. This method calculates the theoretical dosage based on a simplified precipitation reaction formula and uses PID algorithms to adjust the dosage based on the monitoring deviation of the effluent phosphorus concentration. However, this method struggles to respond promptly to sudden changes in influent load, easily leading to instantaneous exceedances of standards in the effluent or wasted chemicals. Second, there are control methods based on macroscopic efficiency models and multi-objective optimization. This method correlates phosphorus removal efficiency with parameters such as dosage ratio and temperature through empirical functions and integrates multiple indicators for optimization decisions. However, this method has limitations in predictive accuracy and adaptability to complex operating conditions. Therefore, existing technologies face widespread challenges in addressing water quality fluctuations and achieving precise, forward-looking control due to the aforementioned lag or insufficient mechanistic descriptions.

[0003] Therefore, there is an urgent need for a method and system for adding phosphorus removal agents for chemical phosphorus removal in wastewater treatment plants. Summary of the Invention

[0004] (a) Technical problems to be solved

[0005] In view of the above-mentioned shortcomings and deficiencies of the prior art, this application provides a method and system for adding phosphorus removal agents for chemical phosphorus removal in wastewater treatment plants, which solves the defects of the prior art chemical phosphorus removal control technology, such as simplified model mechanism, control lag, lack of predictive ability and insufficient adaptability.

[0006] (II) Technical Solution

[0007] To achieve the above objectives, the main technical solutions adopted in this application include:

[0008] In a first aspect, embodiments of this application provide a method for adding phosphorus removal agents for chemical phosphorus removal in wastewater treatment plants, including:

[0009] Real-time collection of operating parameters of the wastewater treatment plant, including: influent flow rate, influent phosphate concentration, pH value, temperature, phosphorus removal agent dosage, mixing intensity characterization parameters, and effluent phosphorus concentration;

[0010] The collected operating parameters are input into a pre-built dynamic physicochemical model to simulate the kinetic changes of the phosphorus removal process and to predict the trajectory of effluent phosphorus concentration changes within a set prediction time domain.

[0011] The dynamic physicochemical model includes: a precipitation module for describing the kinetic process of the hydrolysis of phosphorus removal agents to form hydrated metal oxide precipitates; an aging module for describing the kinetic process of the hydrated metal oxides aging and deactivating over time; and an adsorption module for describing the kinetic process of phosphate adsorption or co-precipitation on the surface of the hydrated metal oxides.

[0012] Based on the trajectory of effluent phosphorus concentration change within the future predicted time domain, a model predictive control algorithm is adopted to obtain the optimal phosphorus removal agent dosage sequence for the future predicted time domain with the optimization objective of ensuring stable effluent phosphorus concentration and minimizing phosphorus removal agent dosage within the future predicted time domain.

[0013] Optionally, in some embodiments of this application, the dynamic physicochemical model further includes a chemical equilibrium module;

[0014] The chemical equilibrium module is used to describe the acid-base balance of carbonic acid and phosphoric acid, to calculate the speciation distribution of various phosphate ions in the solution under different pH values ​​and temperature conditions, and to provide the speciation distribution results to the adsorption module.

[0015] Optionally, in some embodiments of this application, the step of inputting the collected operating parameters into a pre-constructed dynamic physicochemical model to simulate the kinetic changes of the phosphorus removal process and to predict the trajectory of effluent phosphorus concentration changes within a set prediction time domain in the future specifically includes:

[0016] Using the operating parameters as initial conditions, the calculations are recursively performed at each time point within the future prediction time domain. The complete calculations of the dynamic physicochemical model are repeated at each prediction time point to obtain the predicted values ​​of effluent phosphorus concentration at each time point within the future prediction time domain, thus forming the trajectory of effluent phosphorus concentration change.

[0017] Optionally, in some embodiments of this application, the dynamic physicochemical model performing a complete calculation specifically includes:

[0018] In the precipitation module, the amount of newly generated highly active hydrated metal oxides is obtained by using the precipitation reaction kinetic equation based on the current dosage of phosphorus removal agent, pH value and temperature.

[0019] In the aging module, based on the amount of highly active hydrated metal oxides, the mixing intensity characterization parameters, and the state of adsorbed phosphorus, the real-time conversion rate between highly active, low-active, and inactive hydrated metal oxides is obtained through the aging kinetic equation set, and the number of remaining active sites on the surface of the hydrated metal oxides is updated.

[0020] In the chemical equilibrium module, the speciation coefficients of various phosphate ions are obtained based on the pH value and temperature.

[0021] In the adsorption module, based on the current phosphate concentration of the solution, the number of remaining active sites on the surface of the hydrated metal oxide, and the speciation coefficient of phosphate ions, the real-time removal amount of phosphate adsorbed on the surface and encapsulated internally is obtained through the adsorption and co-precipitation kinetic equation.

[0022] Based on the influent phosphate concentration, the real-time removal amount of phosphate adsorbed on the surface and encapsulated internally, and the reactor volume, the predicted value of the effluent phosphorus concentration at the current predicted time point is obtained through the material balance differential equation.

[0023] Optionally, in some embodiments of this application, the method further includes: adaptively calibrating the dynamic physicochemical model, specifically including:

[0024] Based on real-time acquired temperature data, the reaction rate constant in the dynamic physicochemical model is corrected for temperature using the Arrhenius equation, and the corrected reaction rate constant is then updated in the dynamic physicochemical model.

[0025] Historical and current operating parameters are periodically input into the optimization algorithm, which then updates at least one key parameter in the dynamic physicochemical model online. The key parameter includes the precipitation rate coefficient, aging rate coefficient, and adsorption equilibrium constant.

[0026] Based on the updated key parameters, the internal variables of the dynamic physicochemical model are reinitialized to obtain the calibrated dynamic physicochemical model.

[0027] Optionally, in some embodiments of this application, the goal of the optimization algorithm is to minimize the mean square error between the predicted value of effluent phosphorus concentration at the predicted time point obtained by the dynamic physicochemical model and the actual detected value.

[0028] The optimization algorithm is one of the following: recursive least squares method, extended Kalman filter, particle filter, or gradient descent method.

[0029] Optionally, in some embodiments of this application, the mixing intensity characterization parameter is the average velocity gradient G value; the G value is calculated by real-time acquisition of the stirrer frequency, stirrer blade characteristic parameters, and reaction tank geometry, and the calculation formula is:

[0030] ;

[0031] Where P is the stirring power. V is the fluid dynamic viscosity, and V is the volume of the reaction tank.

[0032] Optionally, in some embodiments of this application, the method further includes, before inputting the collected operating parameters into the dynamic physicochemical model:

[0033] The collected operating parameters are cleaned, outliers are removed, and missing values ​​are filled in;

[0034] Based on the concentration of the phosphorus removal agent, the initial concentrations of highly active hydrated metal oxides and low-activity hydrated metal oxides are obtained and used as the initial state variables of the dynamic physicochemical model.

[0035] Optionally, in some embodiments of this application, the step of employing a model predictive control algorithm to obtain the optimal phosphorus removal agent dosage sequence for the future predicted time domain, with the optimization objective of ensuring stable effluent phosphorus concentration and minimizing phosphorus removal agent dosage within a future set prediction time domain, specifically includes:

[0036] At the beginning of each future prediction time domain, using the current operating parameters as initial conditions, the trajectory of effluent phosphorus concentration change within the future prediction test domain is obtained as the prediction input for the optimization problem. An optimization problem with constraints is constructed and solved. The decision variables of the optimization problem are the phosphorus removal agent addition sequence within the future prediction time domain. The objective function is the weighted sum of the squares of the deviations between the predicted effluent phosphorus concentration and the target set value at each time point within the future prediction time domain, plus the penalty terms for the phosphorus removal agent addition amount and the rate of change of the phosphorus removal agent addition amount. The constraints are the dosing pump flow limit and the upper limit of effluent phosphorus concentration.

[0037] The optimization problem is solved using a numerical optimization algorithm to obtain the optimal phosphorus removal agent dosage sequence within the future predicted time domain.

[0038] Secondly, embodiments of this application provide a phosphorus removal agent dosing system for chemical phosphorus removal in wastewater treatment plants, comprising:

[0039] The data acquisition layer is used to collect the operating parameters of the wastewater treatment plant in real time. The operating parameters include: influent flow rate, influent phosphate concentration, pH value, temperature, phosphorus removal agent dosage, mixing intensity characterization parameters, and effluent phosphorus concentration.

[0040] The data processing layer is used to input the collected operating parameters into a pre-built dynamic physicochemical model to simulate the kinetic changes of the phosphorus removal process and to predict the trajectory of effluent phosphorus concentration changes within a set prediction time domain.

[0041] The dynamic physicochemical model includes: a precipitation module for describing the kinetic process of the hydrolysis of phosphorus removal agents to form hydrated metal oxide precipitates; an aging module for describing the kinetic process of the hydrated metal oxides aging and deactivating over time; and an adsorption module for describing the kinetic process of phosphate adsorption or co-precipitation on the surface of the hydrated metal oxides.

[0042] The predictive control layer is used to obtain the optimal phosphorus removal agent dosage sequence for the future predicted time domain by adopting a model predictive control algorithm based on the trajectory of effluent phosphorus concentration change within the future predicted time domain, with the optimization objective of ensuring stable effluent phosphorus concentration and minimizing phosphorus removal agent dosage within the future predicted time domain.

[0043] (III) Beneficial Effects

[0044] The beneficial effects of this application are as follows: The phosphorus removal agent dosing method for chemical phosphorus removal in wastewater treatment plants adopts a combination of dynamic physicochemical model and model predictive control algorithm. This model accurately simulates the kinetic changes of the phosphorus removal agent throughout the entire process from hydrolysis and aging to adsorption and phosphorus removal through sedimentation, aging, and adsorption modules, and based on this, accurately predicts the future trajectory of effluent phosphorus concentration. Subsequently, the model predictive control algorithm is used to solve for the optimal dosing sequence with the optimization objective of ensuring stable effluent compliance and minimizing agent dosage in the future time domain. Compared with existing technologies that typically rely on static empirical formulas or simple proportional control, this application can significantly improve the control accuracy and robustness of the phosphorus removal process, effectively addressing the interference caused by fluctuations in influent water quality and changes in agent activity. It not only ensures stable effluent phosphorus concentration compliance while minimizing phosphorus removal agent consumption and saving operating costs, but also reduces the amount of chemical sludge generated, achieving efficient, economical, and stable operation of the wastewater treatment plant. Attached Figure Description

[0045] Figure 1 This is a schematic flowchart of a phosphorus removal agent dosing method for chemical phosphorus removal in a wastewater treatment plant according to an embodiment of this application;

[0046] Figure 2 This is a flowchart illustrating a method for adding phosphorus removal agents for chemical phosphorus removal in a wastewater treatment plant, according to an embodiment of this application.

[0047] Figure 3 This is an internal formula diagram of a dynamic physicochemical model of a phosphorus removal agent dosing method for chemical phosphorus removal in a wastewater treatment plant according to an embodiment of this application;

[0048] Figure 4 This is a schematic diagram illustrating the beneficial effects of a phosphorus removal agent dosing method for chemical phosphorus removal in a wastewater treatment plant according to an embodiment of this application;

[0049] Figure 5 This is a schematic diagram of a phosphorus removal agent dosing system for chemical phosphorus removal in a wastewater treatment plant, according to an embodiment of this application. Detailed Implementation

[0050] To better explain and facilitate understanding of this application, the following detailed description of the application is provided in conjunction with the accompanying drawings and specific embodiments.

[0051] In related technologies, typical chemical phosphorus removal control methods for the dosing of phosphorus removal agents used in wastewater treatment plants mainly include:

[0052] The first type is a control method based on stoichiometry and feedback. A Chinese patent application, "Intelligent Chemical Phosphorus Removal System for Urban Wastewater Treatment Plants" (authorization announcement number: CN117037926B), discloses such a method. This method first calculates the basic PAC (polyaluminum chloride) dosage based on a simplified stoichiometric equation, the theoretical dosage ratio, and a set dosage coefficient. The dosage coefficient is typically set within a fixed range of 1.5-2, or dynamically adjusted using an empirical formula based on the oxidation-reduction potential (ORP) of the anaerobic zone and the dissolved oxygen (DO) of the aerobic zone. The system uses the deviation between the online detection value of the effluent orthophosphate concentration and the set target value as the core feedback. Through PID or similar control algorithms, the basic dosage is corrected in real time, thereby adjusting the dosing pump frequency. This method attempts to combine biological phosphorus removal indicator parameters (ORP, DO) to adjust the redundancy of chemical dosing, but its core remains a combination of a static stoichiometric model and a single feedback loop.

[0053] The second category is a control method based on models and multi-objective optimization. A Chinese patent application, "Intelligent Control Method, Device, Equipment and Storage Medium for Phosphorus Removal Process in Wastewater Treatment Plants" (Authorization Announcement No.: CN119977010B), discloses another approach. This method determines the phosphorus removal agent dosage scheme from pre-stored basic data of the wastewater treatment plant and divides the dosage cycle into multiple control cycles. It determines the agent dosage ratio for each cycle through a target flocculation efficiency calculation formula, in which the chemical phosphorus removal efficiency is expressed as a function of the agent dosage ratio, temperature, and pH. Simultaneously, the method calculates multiple indicators such as agent consumption reduction, sludge production reduction, cost reduction, carbon emission reduction, and stability, and uses the analytic hierarchy process (AHP) to weight these indicators to comprehensively evaluate and verify the merits of the dosage scheme. This method introduces multi-objective optimization and a more refined efficiency function concept; however, its chemical phosphorus removal efficiency function remains a macroscopic, empirical mapping relationship, without deeply revealing its underlying physicochemical kinetic mechanism.

[0054] However, both of the above methods have obvious drawbacks:

[0055] 1. Significant time lag, making it difficult to cope with water quality and quantity shocks: Existing technologies generally employ feedback control of effluent phosphorus concentration. Due to the significant time lag (usually tens of minutes to several hours) between reagent addition, hydrolysis and precipitation, adsorption reaction and effluent monitoring, feedback control can only remedy after an excessive level has already occurred, and cannot predict or suppress it in advance. When the influent total phosphorus load suddenly increases, it easily leads to excessive effluent levels; while when the load decreases, it easily leads to excessive reagent addition.

[0056] 2. Neglecting Reaction Kinetics and Mechanisms: As exemplified by CN117037926B, its core assumptions are based on simple chemical precipitation reactions, relying primarily on a fixed or empirically adjusted dosing coefficient to cover complex factors such as incomplete reactions, competitive adsorption, and aging deactivation. However, in actual wastewater treatment within the pH range (4.5-8), the amount of metal phosphate precipitation is negligible; phosphorus removal mainly occurs through the adsorption and co-precipitation of newly formed hydrated metal oxides. This process involves a series of rapid and interconnected kinetic reactions, including precipitate formation, aging (leading to adsorption site deactivation), and surface complexation. Existing static stoichiometric models cannot describe these dynamic processes, resulting in inaccurate estimations of phosphorus removal capacity, a lack of scientific basis for setting the dosing coefficient, and susceptibility to failure due to changes in operating conditions.

[0057] 3. Poor adaptability to changes in key process conditions (such as mixing intensity, pH, and precipitate age): Phosphorus removal efficiency is highly dependent on mixing intensity (affecting precipitate aging rate and flocculation effect), pH value (affecting surface complexation equilibrium and adsorption morphology), and the accumulation and aging state of added reagents in the system. Although existing methods partially consider pH, they mostly use simple correction factors and do not incorporate dynamic states such as mixing intensity and precipitate aging into the core control model, resulting in poor system adaptability to fluctuations in process conditions and weak control robustness.

[0058] Therefore, there is an urgent need for an intelligent method that can accurately predict the phosphorus removal process in real time and achieve proactive control of reagent dosing.

[0059] To overcome the aforementioned problems, this application proposes an intelligent control method and system for chemical phosphorus removal that integrates a dynamic physicochemical model based on physicochemical mechanisms with adaptive calibration and predictive control. This system enables proactive prediction, precise decision-making, and adaptive adjustment. The aim is to ensure stable compliance of total phosphorus levels in the effluent while minimizing reagent consumption, reducing sludge production, and lowering operating costs.

[0060] To better understand the above technical solutions, exemplary embodiments of this application will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of this application are shown in the drawings, it should be understood that this application can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this application can be understood more clearly and thoroughly, and that the scope of this application can be fully conveyed to those skilled in the art.

[0061] Figure 1 and Figure 2 This is a schematic flowchart illustrating a method for adding phosphorus removal agents for chemical phosphorus removal in a wastewater treatment plant according to an embodiment of this application. Figure 1 and Figure 2 As shown, the method for adding the phosphorus removal agent includes:

[0062] Real-time collection of operating parameters of the wastewater treatment plant, including: influent flow rate, influent phosphate concentration, pH value, temperature, phosphorus removal agent dosage, mixing intensity characterization parameters, and effluent phosphorus concentration;

[0063] In the specific implementation process, the above operating parameters are obtained in real time or periodically through instruments such as influent and effluent flow rate, return flow rate, sludge discharge flow rate, orthophosphate concentration, pH value, temperature, stirring frequency, phosphorus removal agent dosing flow rate, and effluent orthophosphate or total phosphorus concentration.

[0064] In addition to the data mentioned above, the operating parameters also include the geometric dimensions of the reaction tank, the frequency of the agitator, the characteristic parameters of the agitator blades, the reflux flow rate, and the sludge discharge flow rate.

[0065] The mixing intensity is characterized by the average velocity gradient G value. The G value is calculated using real-time data collected from the agitator frequency, impeller characteristics, and reaction tank geometry. The calculation formula is as follows:

[0066] ;

[0067] Where P is the stirring power. V is the fluid dynamic viscosity, and V is the volume of the reaction tank.

[0068] In addition, before inputting the collected operating parameters into the dynamic physicochemical model, the following steps are also included:

[0069] The collected operating parameters are cleaned, outliers are removed, and missing values ​​are filled in;

[0070] Based on the concentration of the phosphorus removal agent, the initial concentrations of highly active hydrated metal oxides and low-activity hydrated metal oxides were obtained and used as the initial state variables for the dynamic physicochemical model.

[0071] Specifically, because online instruments are susceptible to interference from the field environment and the model has high requirements for data quality, the raw data must be rigorously preprocessed before the collected operating parameters are input into the dynamic physicochemical model to ensure the stability and accuracy of the model operation. For data cleaning, statistical methods such as the 3σ principle or box plot method are used to identify and remove pulse-type outliers caused by instrument malfunctions or communication anomalies. Median filtering or low-pass filtering algorithms are used to smooth high-frequency fluctuations caused by measurement noise. For missing value handling, for transient data loss caused by instrument flushing, brief communication interruptions, time-series-based interpolation methods such as linear interpolation or cubic spline interpolation are used to fill the missing data. For long-term data interruptions, model prediction must be paused and an alarm triggered. Furthermore, input parameters with synchronous dimensions need to be normalized to eliminate dimensional influences and improve the convergence speed of the model solution.

[0072] After data preprocessing, the initial state variables of the dynamic physicochemical model need to be set. Specifically, based on the concentration of the reagent in the mixing tank and its chemical properties, such as the content of iron or aluminum salts, the initial concentration of highly active hydrated metal oxides formed instantaneously upon entering the reaction tank is calculated using a preset stoichiometric relationship. This substance has an extremely high specific surface area and is the main carrier for phosphorus adsorption. Simultaneously, based on the amount of partially deactivated precipitate already present in the reaction tank, the initial concentration of less active hydrated metal oxides is estimated. This substance originates from the aging of highly active substances and has a weaker adsorption capacity. These two initial concentrations will serve as the initial state variables for the aging module and the adsorption module in the model, jointly determining the phosphorus removal potential at the initial moment.

[0073] Through the aforementioned data acquisition and preprocessing, and by introducing the average velocity gradient G value as a key characterizing parameter for mixing intensity for real-time calculation, multi-dimensional process information affecting the kinetics of chemical phosphorus removal can be comprehensively and accurately captured, providing high-fidelity input data for the dynamic physicochemical model. Simultaneously, rigorous cleaning, interpolation, and normalization of the raw data effectively eliminates the impact of on-site instrument interference and data noise on model stability. Furthermore, the initial state variables of the model are precisely set based on reagent characteristics and reaction tank inventory. This series of operations ensures the reliability of the model simulation throughout the entire process, from the accuracy of the "source" data to the rationality of the "initial state," significantly improving the accuracy of subsequent dynamic simulation and predictive control. This lays a solid foundation for achieving precise, on-demand dosing of reagents under complex water quality fluctuations.

[0074] After the above processing, the collected operating parameters are input into the pre-constructed dynamic physicochemical model to simulate the kinetic changes of the phosphorus removal process and to predict the trajectory of effluent phosphorus concentration changes within the set prediction time domain in the future.

[0075] The dynamic physicochemical model includes: a precipitation module to describe the kinetic process of the hydrolysis of phosphorus removal agents to form hydrated metal oxide precipitates; an aging module to describe the kinetic process of the aging and deactivation of hydrated metal oxides over time; and an adsorption module to describe the kinetic process of the adsorption or co-precipitation of phosphates on the surface of hydrated metal oxides.

[0076] Furthermore, the dynamic physicochemical model also includes a chemical equilibrium module;

[0077] The chemical equilibrium module is used to describe the acid-base balance of carbonic acid and phosphoric acid, calculate the speciation distribution of various phosphate ions in the solution under different pH and temperature conditions, and provide the speciation distribution results to the adsorption module.

[0078] Taking iron salt chemical phosphorus removal as an example, the specific composition and working mechanism of the dynamic physicochemical model are as follows (e.g.) Figure 3 (as shown)

[0079] The precipitation module describes the kinetics of the hydrolysis reaction that occurs after phosphorus removal agents (such as iron or aluminum salts) are added to water (see [link]). Figure 3 (See Tables 1, 2, and 3). This module simulates metal ions (such as Fe) in reagents based on the hydrolysis equilibrium constant of metal salts and the law of mass action. 3+ or AL 3+ The process involves the combination of hydroxyl ions in water to rapidly form amorphous hydrated metal oxide precipitates. The reaction rate is influenced by pH, temperature, and mixing intensity (G). pH determines the morphological distribution of the hydrolytic organisms, temperature affects the reaction rate constant, and mixing intensity influences the nucleation and growth of the precipitate. The module outputs the concentration of highly active hydrated metal oxides. This nascent precipitate possesses a large surface area and abundant surface hydroxyl groups, serving as the primary source of active sites for subsequent phosphorus adsorption and removal.

[0080] Specifically, the sedimentation module includes:

[0081] Activity classification submodule: Based on the mixing intensity characterization parameters and the dosage of phosphorus removal agent, the initial total amount of newly generated hydrated metal oxide is determined and divided into high-activity state and low-activity state, and assigned corresponding active site densities respectively; this submodule outputs the initial concentration of high-activity state, the initial concentration of low-activity state and the total active site density as substrate input for subsequent adsorption reactions.

[0082] Phosphorus speciation selective adsorption submodule: Receives the active site density and solid phase concentration output from the active fractionation submodule, calculates the distribution coefficient of each phosphate speciation based on the current pH and temperature, and solves the adsorption rate of different phosphate speciations on highly active and inactive surfaces by combining bidentate or monodentate adsorption kinetic equations; This submodule outputs the remaining highly active solid phase concentration, inactive solid phase concentration, and phosphorus loading on the solid phase surface after the adsorption reaction, and transfers the solid phase concentration to the coprecipitation and encapsulation submodule.

[0083] Coprecipitation and Encapsulation Submodule: Using the solid phase concentration output by the phosphorus selective adsorption submodule as the initial condition, this module simulates the dynamic process of some surface phosphorus being encapsulated into the crystal lattice to form an inert state during the aging process of hydrated metal oxides. This process consumes some of the highly active solid phase and transforms it into a less active or inert state. Finally, it outputs the updated amount of highly active hydrated metal oxides, the concentration of less active solid phase, and the amount of phosphorus encapsulated in the crystal lattice, realizing the dynamic update of the activity state of the newly generated solid phase by the precipitation module.

[0084] The aging module describes the aging process of hydrated metal oxides in a reaction tank over time, characterized by crystal structure rearrangement, decrease in specific surface area, and loss of surface active sites (see [link]). Figure 3 (See Table 5). This module, based on the Oswald ripening mechanism and the principle of surface energy minimization, classifies precipitates into two categories: highly active amorphous morphology and low-activity aged morphology. It describes the conversion rate from highly active to low-activity substances using first-order kinetic equations or more complex particle size distribution models. The aging rate is affected by temperature, pH, and initial precipitate concentration. Increased temperature accelerates the aging process, while a pH deviation from the isoelectric point promotes particle aggregation and crystallization. The introduction of this module allows the model to accurately reflect the cumulative effect and activity decay of the reagent under continuous dosing and long-term operating conditions, avoiding underdosing due to overestimating the adsorption capacity of the existing reagent.

[0085] The adsorption module describes the kinetics of phosphate adsorption or co-precipitation on the surface of active hydrated metal oxides via coordination exchange. Based on a surface complexation model and Langmuir or Freundlich adsorption isotherms, this module facilitates coordination exchange between phosphate ions in the liquid phase and active hydroxyl groups on the oxide surface, forming monodentate or bidentate surface complexes (see [link to relevant documentation]). Figure 3 (See Table 4). The adsorption rate is affected by the liquid-phase phosphate concentration, active site density, pH value, temperature, and mixing intensity. Among these, the mixing intensity controls the external diffusion rate of adsorption by affecting the liquid film mass transfer resistance. This module calculates the amount of phosphate transferred from the liquid phase to the solid phase in real time, which is the core step in determining the phosphorus concentration in the effluent.

[0086] Furthermore, the chemical equilibrium module calculates the distribution of phosphate ions in the solution under the current pH and temperature conditions by simultaneously applying the ionization equilibrium equations, charge balance equations, and mass conservation equations for both carbonic and phosphoric acid. For example, phosphoric acid, as a tribasic acid, will exist as H3PO4 at different pH values. - H2PO4 - HPO4 2- and PO4 3- H2PO4 exists in four forms, and in the pH range of 6-8 commonly used in wastewater treatment, H2PO4 is present in these forms. - and HPO4 2- This is the primary form. The chemical equilibrium module first calculates the equilibrium concentrations of five inorganic carbons and phosphates in the aqueous phase based on real-time collected pH, alkalinity, and temperature. For example, when pH=7.0 and temperature is 20℃, this module calculates the concentration of H2PO4 in the solution. - It accounts for approximately 61% of the total phosphate, HPO4 2- It accounts for approximately 39%, while PO4 3- The concentration was extremely low and negligible. Subsequently, the chemical equilibrium module transmitted this detailed speciation result to the adsorption module. This is because hydrated metal oxides exhibit selective adsorption of phosphates—typically for divalent HPO4. 2- Adsorption affinity for monovalent H2PO4 - —Based on the received speciation information, the adsorption module assigns different adsorption rate constants and surface complexation constants to different ionic speciations, thereby more accurately calculating the total phosphate adsorption and removal amount. Through this coupling mechanism, the model can keenly capture the impact of phosphate speciation changes caused by pH fluctuations on phosphorus removal efficiency; for example, when pH increases and HPO4... 2- When the proportion increases, the model predicts that the adsorption capacity of a unit of reagent will increase; conversely, when the pH decreases, it predicts that the adsorption efficiency will decrease, thus ensuring that the simulation and prediction under different water quality conditions have high mechanistic authenticity and prediction accuracy.

[0087] Optionally, in the adsorption module of the dynamic physicochemical model in this embodiment, redox potential is also introduced as a dynamic correction factor for the iron-phosphorus ratio coefficient. Specifically, online monitoring of redox potential is added during data acquisition, and it is input between the chemical equilibrium module and the adsorption module. When the redox potential decreases (i.e., the reducing power increases), it indicates that there may be more reducing substances or a local anaerobic environment in the wastewater. At this time, ferric ions are at risk of being reduced to ferrous ions, and the precipitate formed by ferrous ions has a much lower phosphorus adsorption capacity than ferric ions. The model automatically calls a preset correction function based on the redox potential value to dynamically reduce the effective adsorption capacity of iron salts in the adsorption module and correspondingly increase the calculation weight of the target reagent dosage. This innovation effectively solves the problem of overestimation of phosphorus removal efficiency caused by neglecting the change of iron ion valence state in traditional models when the influent carbon source fluctuates or the returned sludge carries reducing substances, thus improving the robustness of the model in complex redox environments.

[0088] In this embodiment, the construction and prediction steps of the above-mentioned dynamic physicochemical model, by integrating the precipitation module, aging module, adsorption module and chemical equilibrium module, construct a mechanism model system that can completely describe the behavior of phosphorus removal agents throughout their entire life cycle from hydrolysis and precipitation, active aging to adsorption and removal. This significantly improves the response capability and prediction accuracy to complex water quality disturbances and changes in the properties of the agents themselves, and provides a solid model foundation for the subsequent precise and proactive control of the agents.

[0089] In addition, specifically, the collected operating parameters are input into a pre-constructed dynamic physicochemical model to simulate the kinetic changes in the phosphorus removal process and to predict the trajectory of effluent phosphorus concentration changes within a set prediction time domain. This includes:

[0090] Using the operating parameters as initial conditions, the calculation is recursively performed at each time point within the future prediction time domain. The complete calculation of the dynamic physicochemical model is repeated at each prediction time point to obtain the predicted value of the effluent phosphorus concentration at each time point within the future prediction time domain, thus forming the trajectory of effluent phosphorus concentration change.

[0091] A complete calculation performed by a dynamic physicochemical model specifically includes:

[0092] In the precipitation module, the amount of newly generated highly active hydrated metal oxides is obtained by using the precipitation reaction kinetic equation based on the current dosage of phosphorus removal agent, pH value and temperature.

[0093] In the aging module, based on the amount of highly active hydrated metal oxides, the mixing intensity characterization parameters, and the state of adsorbed phosphorus, the real-time conversion rate between highly active, low-active, and inactive hydrated metal oxides is obtained through the aging kinetic equation set, and the number of remaining active sites on the surface of hydrated metal oxides is updated.

[0094] In the chemical equilibrium module, the speciation coefficients of various phosphate ions are obtained based on pH value and temperature;

[0095] In the adsorption module, based on the current phosphate concentration of the solution, the number of remaining active sites on the surface of the hydrated metal oxide, and the speciation coefficient of phosphate ions, the real-time removal amount of phosphate adsorbed on the surface and encapsulated inside is obtained through the adsorption and coprecipitation kinetic equation.

[0096] Based on the influent phosphate concentration, the real-time removal rate of phosphate adsorbed on the surface and encapsulated internally, and the reactor volume, the predicted value of the effluent phosphate concentration at the current prediction time point is obtained through the material balance differential equation.

[0097] To illustrate more intuitively how dynamic physicochemical models achieve rolling predictions through recursive calculations, the following explanation will be based on a specific virtual scenario.

[0098] Assuming the current time t=0, a future prediction time domain of 30 minutes is defined and discretized into three prediction time points: t=10 minutes, t=20 minutes, and t=30 minutes. The model uses operating parameters collected at t=0, such as influent flow rate, influent phosphate concentration, pH value, temperature, phosphorus removal agent dosage, and mixing intensity G, as initial conditions, and begins recursively calculating forward along the time axis.

[0099] In the first complete calculation predicting the effluent phosphorus concentration at t=10 minutes, the precipitation module first calculates the amount of newly generated highly active hydrated metal oxides and the number of their surface active sites based on the current reagent dosage, pH value, and temperature using hydrolysis kinetic equations. The aging module then updates the active state of the reagents remaining in the reaction tank, comprehensively considering the conversion rate of existing highly active substances to less active substances, and subtracting the occupation of active sites by already adsorbed phosphorus, to obtain the number of remaining active sites currently available for adsorption. The chemical equilibrium module solves for the ionization equilibrium of phosphoric acid based on pH value and temperature, obtaining H2PO4. - and HPO4 2- The distribution coefficients of different ionic forms are calculated. The adsorption module combines the number of remaining active sites with the form distribution coefficients and calculates the amount of phosphate removed by adsorption using the surface complexation kinetic equation. Finally, the model uses the material balance differential equation to comprehensively consider the phosphorus load brought in by the influent, the amount of phosphorus removed by adsorption, and the reactor volume to calculate the predicted value of the effluent phosphorus concentration at t=10 minutes.

[0100] After completing the prediction at t=10 minutes, the model uses the calculated effluent phosphorus concentration, updated concentrations of highly and less active hydrated metal oxides, and the number of remaining active sites as new initial state variables. Under the assumption of unchanged operating conditions, the model repeats the above calculation process of precipitation, aging, chemical equilibrium, adsorption, and material equilibrium, recursively obtaining the predicted effluent phosphorus concentrations at t=20 minutes and t=30 minutes. Thus, the model outputs a trajectory of effluent phosphorus concentration changes over the next 30 minutes, composed of three prediction points. This trajectory not only reflects the dynamic trend of phosphorus concentration changes but, more importantly, couples microscopic mechanisms such as reagent hydrolysis and precipitation, aging and deactivation, dynamic changes in adsorption sites, and the distribution of phosphate chemical speciation into each recursive calculation step. This provides a highly reliable predictive basis for the subsequent model predictive control algorithm to solve for the optimal reagent dosage.

[0101] This recursive mechanism can effectively capture nonlinear processes such as the decay of reagent activity over time, dynamic saturation of adsorption sites, and changes in adsorption affinity caused by pH fluctuations. It significantly improves the accuracy and timeliness of predicting phosphorus concentration in effluent under complex operating conditions, providing a reliable decision-making basis for subsequent forward-looking optimal control, and ensuring accurate dosing of reagents and effective control of operating costs while meeting effluent standards.

[0102] Finally, the method for adding phosphorus removal agents for chemical phosphorus removal in wastewater treatment plants according to embodiments of this application further includes:

[0103] Based on the trajectory of effluent phosphorus concentration change within the future predicted time domain, a model predictive control algorithm is adopted to obtain the optimal phosphorus removal agent dosage sequence for the future predicted time domain with the optimization objective of ensuring stable effluent phosphorus concentration and minimizing phosphorus removal agent dosage within the future predicted time domain.

[0104] Specifically, at the beginning of each future prediction time domain, using the current operating parameters as initial conditions, the trajectory of effluent phosphorus concentration change within the future prediction test domain is obtained as the prediction input for the optimization problem. An optimization problem with constraints is constructed and solved. The decision variables of the optimization problem are the phosphorus removal agent addition sequence within the future prediction time domain. The objective function is the weighted sum of the squares of the deviations between the predicted effluent phosphorus concentration and the target set value at each time point within the future prediction time domain, the penalty term for the phosphorus removal agent addition amount and the rate of change of the phosphorus removal agent addition amount, and the constraints are the dosing pump flow limit and the upper limit of effluent phosphorus concentration.

[0105] The optimization problem is solved by numerical optimization algorithm to obtain the optimal phosphorus removal agent dosage sequence within the future predicted time domain.

[0106] For example, suppose at the current time t=0, the trajectory of effluent phosphorus concentration change over the next 30 minutes has been predicted, yielding predicted values ​​of 0.65 mg / L (10 minutes), 0.55 mg / L (20 minutes), and 0.48 mg / L (30 minutes) under the current dosage of 50 mg / L. However, while this trajectory shows a decreasing trend in phosphorus concentration, it has not yet been compared and optimized against the target setpoint, and the cost of chemical consumption has not been considered. Therefore, the model predictive control algorithm intervenes, using the trajectory of effluent phosphorus concentration change over the future predicted time domain as the prediction input for the optimization problem, constructing a constrained optimization problem. The decision variables for this optimization problem are the phosphorus removal agent dosing sequence within the future predicted time domain. Assuming the control time domain is consistent with the predicted time domain, it is necessary to determine the agent dosing amount at three time points: 10 minutes, 20 minutes, and 30 minutes in the future. The objective function consists of a weighted sum of three terms: the sum of squares of the deviations between the predicted effluent phosphorus concentration and the target setpoint (assumed to be 0.5 mg / L) at each time point within the future predicted time domain, a penalty term for the phosphorus removal agent dosing amount itself (used to control the cost of the agent), and a penalty term for the rate of change of the phosphorus removal agent dosing amount (used to prevent drastic fluctuations in the dosing amount from impacting the actuator). The constraints include the flow rate limit of the dosing pump (assumed to be a minimum dosing amount of 20 mg / L and a maximum of 100 mg / L) and the upper limit of the effluent phosphorus concentration (assumed to be no more than 0.5 mg / L according to the discharge standard).

[0107] Subsequently, numerical optimization algorithms (such as sequential quadratic programming or interior point methods) were used to solve the above optimization problem. During the solution process, the algorithm, within the constraints, tried different combinations of future dosages and re-input each candidate dosage sequence into the dynamic physicochemical model, continuously predicting the corresponding future effluent phosphorus concentration trajectory, and then calculating the objective function value corresponding to that trajectory. Through iterative optimization, the algorithm finally found an optimal phosphorus removal agent dosage sequence that minimizes the objective function. Assuming the solution is: dosage adjusted to 65 mg / L in the next 10 minutes, 55 mg / L in the next 20 minutes, and 45 mg / L in the next 30 minutes. This sequence means: at the current moment, the agent dosage needs to be immediately increased from 50 mg / L to 65 mg / L to quickly address the current high phosphorus concentration; after 10 minutes, as the phosphorus concentration decreases, the dosage can be appropriately reduced to 55 mg / L; and after another 10 minutes, it can be further reduced to 45 mg / L, thereby minimizing agent consumption while meeting effluent standards. Therefore, only the first control action in the optimal sequence is executed, that is, the current dosage is adjusted to 65 mg / L, and the latest operating parameters are collected again at the next sampling time (e.g., after 10 minutes), the model state is updated, and rolling optimization is performed again, thus achieving continuous rolling time-domain control. Through this mechanism, the model predictive control algorithm can proactively respond to future water quality changes and adjust the dosage of the reagent in advance, avoiding the overshoot or undershoot problems caused by the lag in traditional feedback control. Ultimately, while ensuring that the phosphorus concentration in the effluent is stably up to standard (all predicted points are controlled below 0.5 mg / L), the minimum dosage of phosphorus removal reagent and the smoothness of the dosing process are achieved.

[0108] The phosphorus removal agent dosing method for chemical phosphorus removal in wastewater treatment plants in this embodiment further includes: adaptive calibration of the dynamic physicochemical model, specifically including:

[0109] Based on real-time acquired temperature data, the reaction rate constant in the dynamic physicochemical model is corrected for temperature using the Arrhenius equation, and the corrected reaction rate constant is then updated in the dynamic physicochemical model.

[0110] Historical and current operating parameters are periodically input into the optimization algorithm, which then updates at least one key parameter in the dynamic physicochemical model online. The key parameters include the precipitation rate coefficient, aging rate coefficient, and adsorption equilibrium constant.

[0111] Based on the updated key parameters, the internal variables of the dynamic physicochemical model are reinitialized to obtain the calibrated dynamic physicochemical model.

[0112] The goal of the optimization algorithm is to minimize the mean square error between the predicted phosphorus concentration in the effluent at the predicted time point obtained by the dynamic physicochemical model and the actual detected value.

[0113] The optimization algorithm is one of the following: recursive least squares, extended Kalman filter, particle filter, or gradient descent.

[0114] The Arrhenius equation describes the exponential relationship between reaction rate and temperature. This equation can be used to correct the baseline rate constant of the model at the standard temperature to the actual rate constant at the current temperature in real time, and update the corrected value to the precipitation, aging and adsorption modules, so that the model can automatically adapt to the effects of diurnal temperature difference and seasonal changes on reaction kinetics.

[0115] After completing the online update of key parameters, the model's internal variables are reinitialized based on the updated parameters. This includes recalculating the cumulative concentration of highly reactive hydrated metal oxides, correcting for aging and conversion stockpiles, and updating the saturation state of adsorption sites, thereby obtaining a calibrated dynamic physicochemical model. Through this adaptive calibration mechanism combining temperature compensation and online parameter updates, the model possesses the ability to continuously learn and self-correct, effectively coping with various disturbances and drifts during long-term operation, maintaining high-precision predictive performance, and providing a reliable foundation for subsequent model predictive control.

[0116] In summary, the embodiments of this application integrate a dynamic physicochemical prediction model that combines mechanism and data. For the first time, it introduces the advanced dynamic physicochemical modeling concept describing the precipitation, aging, and adsorption co-precipitation of metal oxides into the real-time control of industrial chemical phosphorus removal. This differs from all existing technologies based on simple stoichiometry or empirical efficiency functions, enabling a more accurate and dynamic characterization of the phosphorus removal process, and is the cornerstone of precise control. Furthermore, this application also employs proactive control based on model predictive control, using the dynamic physicochemical model to predict future effluent concentrations. This allows for proactive and advance adjustment of the dosing strategy, fundamentally overcoming the lag problem of feedback control and effectively addressing influent shocks. Moreover, adaptive calibration is introduced to calibrate the physical parameters in the dynamic physicochemical model online. This maintains the model's interpretability and robustness while enabling it to adapt to different wastewater treatment plants and operating conditions, solving the problem of fixed and uncommon dynamic physicochemical model parameters.

[0117] Therefore, as Figure 4As shown, the embodiments of this application avoid conservative over-addition to compensate for model inaccuracies and lags by using the accuracy and foresight of model predictions. Meanwhile, adaptive parameter calibration ensures that the model operates at its optimal state over the long term, potentially reducing the dosage of chemicals while maintaining compliance with standards. The predictive control algorithm's forward-looking control effectively mitigates fluctuations in effluent phosphorus concentration caused by influent load fluctuations, ensuring that the total phosphorus concentration in the effluent remains more stable and safer below the emission limits, significantly reducing the risk of exceeding standards and minimizing the deviation between effluent values ​​and target values. Furthermore, the precise reduction in chemical sludge dosage directly leads to a decrease in chemical sludge production, saving on sludge treatment and disposal costs and meeting environmental protection requirements for energy conservation and emission reduction. Moreover, by considering dynamic factors such as pH, mixing intensity, and sediment age, and combining adaptive calibration, the control strategy can be automatically adjusted under seasonal changes, water quality fluctuations, and operational mode adjustments, maintaining excellent control performance without frequent manual intervention. Finally, the dynamic physicochemical model in this embodiment can be used to predict the results of its internal state variables, providing operators with in-depth insights into the phosphorus removal reaction process and facilitating more advanced process optimization (such as optimal dosing point selection and mixing condition optimization).

[0118] in addition, Figure 5 This is a schematic diagram of a phosphorus removal agent dosing system for chemical phosphorus removal in a wastewater treatment plant, as described in an embodiment of this application. Figure 5 As shown, the phosphorus removal agent dosing system includes:

[0119] The data acquisition layer is used to collect the operating parameters of the wastewater treatment plant in real time. These parameters include: influent flow rate, influent phosphate concentration, pH value, temperature, phosphorus removal agent dosage, mixing intensity characterization parameters, and effluent phosphorus concentration.

[0120] The data processing layer is used to input the collected operating parameters into a pre-built dynamic physicochemical model to simulate the kinetic changes of the phosphorus removal process and to predict the trajectory of effluent phosphorus concentration changes within a set prediction time domain.

[0121] The dynamic physicochemical model includes: a precipitation module to describe the kinetic process of the hydrolysis of phosphorus removal agents to form hydrated metal oxide precipitates; an aging module to describe the kinetic process of the aging and deactivation of hydrated metal oxides over time; and an adsorption module to describe the kinetic process of the adsorption or co-precipitation of phosphates on the surface of hydrated metal oxides.

[0122] The predictive control layer is used to obtain the optimal phosphorus removal agent dosage sequence for the future predicted time domain by adopting a model predictive control algorithm based on the trajectory of phosphorus concentration change in the effluent within the future predicted time domain. The optimization objective is to ensure that the phosphorus concentration in the effluent reaches the standard stably and that the dosage of phosphorus removal agent is minimized within the future predicted time domain.

[0123] The system also includes:

[0124] The execution control layer obtains the optimal dosage sequence from the predictive control layer and extracts the first control action corresponding to the current moment in the sequence. It converts this action into a standard 4-20mA analog signal or digital pulse signal, which drives the frequency converter or stepper motor of the phosphorus removal agent dosing pump in real time to precisely adjust the dosage flow rate. At the same time, this layer feeds back execution information such as the actual dosage and pump operating status to the human-machine interface in real time, forming a control closed loop. It can also automatically switch to a safe mode in case of communication interruption or abnormality, and maintain the execution of the last valid command.

[0125] The human-machine interface provides a graphical platform for process monitoring and parameter configuration. It dynamically displays key data such as influent water quality, predicted phosphorus concentration changes, current reagent dosage, and actual effluent phosphorus concentration in the form of trend curves and dashboards, allowing operators to intuitively grasp the system's operating status. The interface also allows operators to flexibly set effluent target values, alarm thresholds, and relevant parameters of the predictive control algorithm (such as prediction time domain length and optimization weight coefficients) according to emission standards or process requirements. In emergencies, it can be manually switched to local control mode to ensure system safety and flexibility. Through the close collaboration of these five layers, this system achieves intelligent and precise control of the chemical phosphorus removal process, effectively reducing reagent consumption and operating costs while ensuring stable effluent compliance.

[0126] The phosphorus removal agent dosing system for chemical phosphorus removal in wastewater treatment plants, as described in this application, achieves closed-loop control throughout the entire process—from data perception and mechanism modeling to rolling optimization, precise execution, and visual monitoring—by constructing a five-layer collaborative architecture comprising a data acquisition layer, a data processing layer, a predictive control layer, an execution control layer, and a human-machine interface. This is because the system introduces a dynamic physicochemical model integrating sedimentation, aging, and adsorption mechanisms at the data processing layer, and employs a model predictive control algorithm aimed at minimizing agent consumption at the predictive control layer. This not only allows for accurate prediction of future dynamic trends in effluent phosphorus concentration based on real-time operating conditions, but also enables the pre-determining and execution of the optimal agent dosing sequence while meeting emission constraints, effectively overcoming the large lag and overdosing problems of traditional control methods. Simultaneously, the real-time signal conversion at the execution control layer and the flexible parameter configuration of the human-machine interface ensure the precise implementation of control commands and the visible and controllable operation of the system. Through this complete intelligent control system, the system significantly improves the response speed and control accuracy of the chemical phosphorus removal process, ensuring that the effluent quality meets the standards while minimizing the amount of reagents added, bringing significant economic benefits and operational stability to the wastewater treatment plant.

[0127] In the description of this application, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0128] In this application, unless otherwise expressly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0129] In this application, unless otherwise expressly specified and limited, "above" or "below" the second feature can mean that the first and second features are in direct contact, or that they are in indirect contact through an intermediate medium. Furthermore, "above," "on top of," and "over" the second feature can mean that the first feature is directly above or diagonally above the second feature, or simply that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature can mean that the first feature is directly below or diagonally below the second feature, or simply that the first feature is at a lower horizontal level than the second feature.

[0130] In the description of this specification, the terms "one embodiment," "some embodiments," "embodiment," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0131] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make modifications, alterations, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for adding phosphorus removal agents for chemical phosphorus removal in wastewater treatment plants, characterized in that, include: Real-time collection of operating parameters of the wastewater treatment plant, including: influent flow rate, influent phosphate concentration, pH value, temperature, phosphorus removal agent dosage, mixing intensity characterization parameters, and effluent phosphorus concentration; The collected operating parameters are input into a pre-built dynamic physicochemical model to simulate the kinetic changes of the phosphorus removal process and to predict the trajectory of effluent phosphorus concentration changes within a set prediction time domain. The dynamic physicochemical model includes: a precipitation module for describing the kinetic process of the hydrolysis of phosphorus removal agents to form hydrated metal oxide precipitates; an aging module for describing the kinetic process of the hydrated metal oxides aging and deactivating over time; and an adsorption module for describing the kinetic process of phosphate adsorption or co-precipitation on the surface of the hydrated metal oxides. Based on the trajectory of effluent phosphorus concentration change within the future predicted time domain, a model predictive control algorithm is adopted to obtain the optimal phosphorus removal agent dosage sequence for the future predicted time domain with the optimization objective of ensuring stable effluent phosphorus concentration and minimizing phosphorus removal agent dosage within the future predicted time domain.

2. The method for adding phosphorus removal agents for chemical phosphorus removal in wastewater treatment plants according to claim 1, characterized in that, The dynamic physicochemical model also includes a chemical equilibrium module; The chemical equilibrium module is used to describe the acid-base balance of carbonic acid and phosphoric acid, to calculate the speciation distribution of various phosphate ions in the solution under different pH values ​​and temperature conditions, and to provide the speciation distribution results to the adsorption module.

3. The method for adding phosphorus removal agents for chemical phosphorus removal in wastewater treatment plants according to claim 2, characterized in that, The process involves inputting the collected operating parameters into a pre-constructed dynamic physicochemical model to simulate the kinetic changes in the phosphorus removal process and to predict the trajectory of effluent phosphorus concentration changes within a set prediction time domain. Specifically, this includes: Using the operating parameters as initial conditions, the calculations are recursively performed at each time point within the future prediction time domain. The complete calculations of the dynamic physicochemical model are repeated at each prediction time point to obtain the predicted values ​​of effluent phosphorus concentration at each time point within the future prediction time domain, thus forming the trajectory of effluent phosphorus concentration change.

4. The method for adding phosphorus removal agents for chemical phosphorus removal in wastewater treatment plants according to claim 3, characterized in that, The dynamic physicochemical model performs a complete calculation, which specifically includes: In the precipitation module, the amount of newly generated highly active hydrated metal oxides is obtained by using the precipitation reaction kinetic equation based on the current dosage of phosphorus removal agent, pH value and temperature. In the aging module, based on the amount of highly active hydrated metal oxides, the mixing intensity characterization parameters, and the state of adsorbed phosphorus, the real-time conversion rate between highly active, low-active, and inactive hydrated metal oxides is obtained through the aging kinetic equation set, and the number of remaining active sites on the surface of the hydrated metal oxides is updated. In the chemical equilibrium module, the speciation coefficients of various phosphate ions are obtained based on the pH value and temperature. In the adsorption module, based on the current phosphate concentration of the solution, the number of remaining active sites on the surface of the hydrated metal oxide, and the speciation coefficient of phosphate ions, the real-time removal amount of phosphate adsorbed on the surface and encapsulated internally is obtained through the adsorption and co-precipitation kinetic equation. Based on the influent phosphate concentration, the real-time removal amount of phosphate adsorbed on the surface and encapsulated internally, and the reactor volume, the predicted value of the effluent phosphorus concentration at the current predicted time point is obtained through the material balance differential equation.

5. The method for adding phosphorus removal agents for chemical phosphorus removal in wastewater treatment plants according to claim 1, characterized in that, The method further includes: adaptive calibration of the dynamic physicochemical model, specifically including: Based on real-time acquired temperature data, the reaction rate constant in the dynamic physicochemical model is corrected for temperature using the Arrhenius equation, and the corrected reaction rate constant is then updated in the dynamic physicochemical model. Historical and current operating parameters are periodically input into the optimization algorithm, which then updates at least one key parameter in the dynamic physicochemical model online. The key parameter includes the precipitation rate coefficient, aging rate coefficient, and adsorption equilibrium constant. Based on the updated key parameters, the internal variables of the dynamic physicochemical model are reinitialized to obtain the calibrated dynamic physicochemical model.

6. The method for adding phosphorus removal agents for chemical phosphorus removal in wastewater treatment plants according to claim 4, characterized in that, The goal of the optimization algorithm is to minimize the mean square error between the predicted value of effluent phosphorus concentration at the predicted time point obtained by the dynamic physicochemical model and the actual detected value. The optimization algorithm is one of the following: recursive least squares method, extended Kalman filter, particle filter, or gradient descent method.

7. The method for adding phosphorus removal agents for chemical phosphorus removal in wastewater treatment plants according to claim 1, characterized in that, The mixing intensity characterization parameter is the average velocity gradient G value; the G value is calculated by real-time acquisition of the stirrer frequency, stirrer blade characteristic parameters, and reaction tank geometry, and the calculation formula is: ； Where P is the stirring power. V is the fluid dynamic viscosity, and V is the volume of the reaction tank.

8. The method for adding phosphorus removal agents for chemical phosphorus removal in wastewater treatment plants according to claim 1, characterized in that, Before inputting the collected operating parameters into the dynamic physicochemical model, the method further includes: The collected operating parameters are cleaned, outliers are removed, and missing values ​​are filled in; Based on the concentration of the phosphorus removal agent, the initial concentrations of highly active hydrated metal oxides and low-activity hydrated metal oxides are obtained and used as the initial state variables of the dynamic physicochemical model.

9. The method for adding phosphorus removal agents for chemical phosphorus removal in wastewater treatment plants according to claim 1, characterized in that, The model predictive control algorithm, with the optimization objective of ensuring stable effluent phosphorus concentration and minimizing phosphorus removal agent dosage within a future predicted time domain, specifically obtains the optimal phosphorus removal agent dosage sequence for the future predicted time domain, including: At the beginning of each future prediction time domain, using the current operating parameters as initial conditions, the trajectory of effluent phosphorus concentration change within the future prediction test domain is obtained as the prediction input for the optimization problem. An optimization problem with constraints is constructed and solved. The decision variables of the optimization problem are the phosphorus removal agent addition sequence within the future prediction time domain. The objective function is the weighted sum of the squares of the deviations between the predicted effluent phosphorus concentration and the target set value at each time point within the future prediction time domain, plus the penalty terms for the phosphorus removal agent addition amount and the rate of change of the phosphorus removal agent addition amount. The constraints are the dosing pump flow limit and the upper limit of effluent phosphorus concentration. The optimization problem is solved using a numerical optimization algorithm to obtain the optimal phosphorus removal agent dosage sequence within the future predicted time domain.

10. A phosphorus removal agent dosing system for chemical phosphorus removal in wastewater treatment plants, characterized in that, include: The data acquisition layer is used to collect the operating parameters of the wastewater treatment plant in real time. The operating parameters include: influent flow rate, influent phosphate concentration, pH value, temperature, phosphorus removal agent dosage, mixing intensity characterization parameters, and effluent phosphorus concentration. The data processing layer is used to input the collected operating parameters into a pre-built dynamic physicochemical model to simulate the kinetic changes of the phosphorus removal process and to predict the trajectory of effluent phosphorus concentration changes within a set prediction time domain. The dynamic physicochemical model includes: a precipitation module for describing the kinetic process of the hydrolysis of phosphorus removal agents to form hydrated metal oxide precipitates; an aging module for describing the kinetic process of the hydrated metal oxides aging and deactivating over time; and an adsorption module for describing the kinetic process of phosphate adsorption or co-precipitation on the surface of the hydrated metal oxides. The predictive control layer is used to obtain the optimal phosphorus removal agent dosage sequence for the future predicted time domain by adopting a model predictive control algorithm based on the trajectory of effluent phosphorus concentration change within the future predicted time domain, with the optimization objective of ensuring stable effluent phosphorus concentration and minimizing phosphorus removal agent dosage within the future predicted time domain.

Images