Method and system for dynamically determining the dosage of a medicament for a dense medium high-density settling tank

Abstract

The present application relates to the technical field of water treatment, and provides a method and system for dynamically determining the medicament adding amount of a heavy medium high-density sedimentation tank, which comprises the following steps: taking the initial PAC adding concentration as a benchmark, combining the turbidity treatment amount change rate, the influent flow rate change rate, the treatment load change rate, the reflux ratio change rate and the environmental correction coefficient to dynamically determine the PAC adding concentration at the current time; outputting the corresponding PAC adding amount control instruction according to the PAC adding concentration; taking the initial PAM adding concentration as a benchmark, combining the turbidity treatment amount change rate, the influent flow rate change rate, the PAC adding concentration change rate, the magnetite sand adding concentration change rate and the PAM environmental correction coefficient to determine the PAM adding concentration at the current time, so that the PAM adding amount is dynamically matched with the PAC adding amount, the magnetite sand adding amount and the working condition fluctuation, the comprehensive medicament cost is further reduced, and the stability of the effluent water quality is improved.

Description

Technical Field

[0001] This invention relates to the field of water treatment technology, and in particular to a method and system for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank. Background Technology

[0002] Heavy media high-density sedimentation tanks are key units in the field of water treatment, and the dosage of PAC (polyaluminum chloride) coagulant directly determines the effluent quality and operating costs. In existing technologies, the PAC dosage is usually adjusted manually based on static empirical formulas or the proportion of influent turbidity.

[0003] However, static control methods cannot respond to dynamic changes in operating conditions, leading to a mismatch between PAC dosage and real-time requirements, resulting in waste of reagents or effluent quality exceeding standards. Meanwhile, PAM (polyacrylamide), as a coagulant aid, is typically added at a roughly fixed ratio based on the PAC dosage, preventing the full realization of the synergistic effect between PAC and PAM. This not only affects flocculation and sedimentation efficiency but also further increases reagent consumption and operating costs. Summary of the Invention

[0004] This invention provides a method and system for dynamically determining the dosage of chemicals in a heavy media high-density sedimentation tank. This addresses the shortcomings of existing static control methods, which cannot respond to dynamic changes in operating conditions, leading to a mismatch between PAC dosage and real-time requirements, resulting in chemical waste or effluent quality exceeding standards. Furthermore, PAM, as a coagulant aid, fails to fully realize the synergistic effect between PAC and PAM, which not only affects the flocculation and sedimentation effect but also further increases chemical consumption and operating costs.

[0005] This invention provides a method for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank, comprising: Collect the influent turbidity, effluent turbidity, influent flow rate, and sludge return ratio of the heavy media high-density sedimentation tank; Based on the dynamic change of the difference between the influent turbidity and the effluent turbidity, a turbidity treatment capacity change rate is generated; based on the real-time fluctuation of the influent flow rate, an influent flow rate change rate is generated; based on the deviation between the turbidity treatment capacity and the standard operating condition treatment load, a treatment load change rate is generated; and based on the real-time change of the sludge return ratio, a return ratio change rate is generated. Based on the initial PAC dosage concentration, and combined with the turbidity treatment rate change rate, influent flow rate change rate, treatment load change rate, reflux ratio change rate, and environmental correction coefficient, the current PAC dosage concentration is dynamically determined; wherein, the environmental correction coefficient is jointly determined by the pH correction coefficient, temperature correction coefficient, conductivity correction coefficient, and safety correction coefficient. According to the PAC dosage concentration, output the corresponding PAC dosage control command; The PAM dosage concentration change rate and the magnetite sand dosage change rate are acquired in real time. Based on the initial PAM dosage concentration, the PAM dosage concentration at the current moment is dynamically determined by combining the turbidity treatment rate change, influent flow rate change, PAC dosage concentration change, magnetite sand dosage concentration change, and PAM environmental correction coefficient.

[0006] According to the present invention, a method for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank includes determining the pH correction coefficient based on the deviation of the real-time pH value from the preset optimal pH range and the pH change rate. Determining the temperature correction coefficient includes: determining the temperature correction coefficient based on the deviation of the real-time temperature from the preset optimal temperature range and the rate of temperature change, and triggering external temperature control measures when the real-time temperature exceeds the preset safe temperature threshold. Determining the conductivity correction coefficient includes: calculating the conductivity correction coefficient based on the deviation between the real-time conductivity and the preset optimal conductivity range and the conductivity change rate, and triggering electrolyte addition or dilution measures when the real-time conductivity exceeds the preset safe conductivity threshold; The safety correction factor is determined by weighted summation based on the fluctuation range of influent flow rate, the fluctuation range of influent water quality, the error of turbidity detection instrument, and the purity deviation of PAC reagent.

[0007] According to the present invention, a method for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank, after outputting the corresponding PAC dosage control command, further includes: Real-time acquisition of the change rate of PAC dosage concentration and the change rate of magnetite sand dosage concentration; Based on the initial PAM dosage concentration, and combined with the turbidity treatment rate change rate, influent flow rate change rate, PAC dosage concentration change rate, magnetite sand dosage concentration change rate, and PAM environmental correction coefficient, the current PAM dosage concentration is dynamically determined.

[0008] According to the present invention, a method for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank includes determining the PAM environmental correction coefficient, comprising: Using an exponential function, based on the deviation of real-time pH from the optimal pH center of PAM, the deviation of real-time temperature from the optimal temperature range of PAM, and the deviation of real-time conductivity from the optimal conductivity range of PAM, and combined with the molecular chain response hysteresis time, the pH correction coefficient, temperature correction coefficient, and conductivity correction coefficient of PAM are determined respectively, and used as the PAM environmental correction coefficient. After determining the PAM environmental correction coefficient, the method further includes controlling the PAM dosage concentration to not exceed the preset upper limit of the PAC dosage concentration.

[0009] According to the present invention, a method for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank, after dynamically determining the PAM dosage concentration at the current moment, further includes: Determine whether the PAC and PAM dosage concentrations exceed the corresponding preset hard constraint boundaries, and determine the total cost rate in real time. The total cost rate includes at least one of the PAC cost rate, PAM cost rate, magnetite sand cost rate, and effluent quality exceeding penalty cost rate. When the concentration of PAC and / or the concentration of PAM exceed the corresponding preset hard constraint boundary, an alarm is triggered and the dosage adjustment range is limited. When the total cost rate exceeds the preset cost limit, the cost optimization mode is activated, and the dosage concentration is limited to increase while ensuring that the water quality meets the standards.

[0010] According to the present invention, a method for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank, after determining the penalty cost rate for exceeding the effluent quality standard, further includes: When the turbidity of the effluent exceeds the first turbidity threshold or the concentration of suspended solids in the effluent exceeds the first concentration threshold, the rate of change of the penalty cost rate for exceeding the effluent quality standard increases dynamically in proportion to the rate of change of the extent of exceeding the standard.

[0011] According to the present invention, a method for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank is provided, wherein the start-up cost optimization mode includes: Compare the current total cost rate with the historical minimum cost rate. If the current total cost rate is lower than the historical minimum cost rate and the ratio of the two is less than a preset threshold, trigger a breakthrough reward. The breakthrough reward is determined according to the logarithmic function of the ratio of the current total cost rate to the historical minimum cost rate, and the historical minimum cost rate is updated based on the result of the logarithmic function.

[0012] According to the present invention, a method for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank, wherein updating the historical minimum cost rate based on the result of the logarithmic function includes: When the breakthrough reward calculated by the logarithmic function exceeds the preset reward threshold, the current working condition feature vector, the current total cost rate and the current addition amount combination are stored as a new optimal record in the memory bank, and the historical minimum cost rate is updated with the current total cost rate. Based on the similarity between the feature vectors of operating conditions, the historical records in the memory are clustered and merged, and historical records that have not been matched for a long time are eliminated.

[0013] According to the present invention, a method for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank, before updating the historical minimum cost rate based on the result of the logarithmic function, further includes: Construct a memory library, which stores historical operating condition feature vectors, historical minimum cost rates, and historical optimal combination of addition amounts; Perform cosine similarity matching between the current working condition feature vector and the historical working conditions in the memory; When a working condition with a similarity higher than a preset threshold is matched, the corresponding historical best dosage combination is used as a reference. When no matching operation condition with similarity higher than the preset threshold is found, the optimal dosage combination is determined by the reinforcement learning model with the goal of minimizing the total cost. After stable operation, the current operation condition characteristics, the corresponding minimum cost rate, and the optimal dosage combination are stored in the memory bank.

[0014] According to the present invention, a method for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank is provided, wherein the reinforcement learning model adopts a reward function, and the reward function includes at least one of a basic reward, a target achievement reward, and a breakthrough reward. The basic reward is negatively correlated with the current total cost rate, and the compliance reward increases as the water quality improves when the effluent quality is better than the preset standard.

[0015] This invention provides a system for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank, comprising: The data acquisition module is used to collect the influent turbidity, effluent turbidity, influent flow rate, and sludge return ratio of the heavy media high-density sedimentation tank. The generation module is used to generate a turbidity treatment capacity change rate based on the dynamic change of the difference between the influent turbidity and the effluent turbidity, to generate an influent flow rate change rate based on the real-time fluctuation of the influent flow rate, to generate a treatment load change rate based on the deviation between the turbidity treatment capacity and the standard operating condition treatment load, and to generate a sludge return ratio change rate based on the real-time change of the sludge return ratio. The first determining module is used to dynamically determine the current PAC dosage concentration based on the initial PAC dosage concentration, combined with the turbidity treatment rate change rate, influent flow rate change rate, treatment load change rate, reflux ratio change rate, and environmental correction coefficient; wherein, the environmental correction coefficient is jointly determined by the pH correction coefficient, temperature correction coefficient, conductivity correction coefficient, and safety correction coefficient. The output module is used to output the corresponding PAC dosage control command according to the PAC dosage concentration; The second determining module is used to acquire the change rate of PAC dosage concentration and the change rate of magnetite sand dosage concentration in real time; based on the initial PAM dosage concentration, combined with the change rate of turbidity treatment volume, the change rate of influent flow rate, the change rate of PAC dosage concentration, the change rate of magnetite sand dosage concentration and the PAM environmental correction coefficient, the PAM dosage concentration at the current moment is dynamically determined.

[0016] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method for dynamically determining the dosage of reagents in a heavy medium high-density sedimentation tank as described above.

[0017] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank as described above.

[0018] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the method for dynamically determining the dosage of reagents in a heavy medium high-density sedimentation tank as described above.

[0019] Beneficial effects The present invention provides a method and system for dynamically determining the dosage of PAC (Potentially Added Acid) in high-density sedimentation tanks using heavy media. This method collects in-water turbidity, effluent turbidity, influent flow rate, and sludge return ratio in real time, and further generates turbidity treatment rate change, influent flow rate change, treatment load change, and return ratio change. These dynamic change rates, along with environmental correction coefficients such as pH, temperature, and conductivity, are used as the basis for determining the PAC dosage. This solves the problem of mismatch between dosage and real-time requirements caused by the inability of static control methods to respond to fluctuations in operating conditions. By basing the dosage calculation on multiple real-time change rates, the PAC dosage can adaptively adjust to the dynamic fluctuations in turbidity treatment load, influent flow rate, and return ratio, significantly improving control accuracy and response speed. This effectively reduces PAC consumption and lowers operating costs while ensuring effluent quality meets standards. By dynamically determining the current PAM dosage concentration, synchronous dynamic matching of PAM dosage with PAC dosage, magnetite sand dosage, and operating condition fluctuations is achieved. This ensures the synergistic adaptation between the coagulation aid effect and the coagulation process, avoids insufficient coagulation aid or excessive waste of reagents due to imbalance in the ratio of the two, further reduces the overall reagent cost, and improves the stability of effluent water quality. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0021] Figure 1 This is a flowchart illustrating the method for dynamically determining the dosage of reagents in a heavy-medium high-density sedimentation tank provided by the present invention.

[0022] Figure 2 This is a schematic diagram of the system for dynamically determining the dosage of reagents in a heavy medium high-density sedimentation tank provided by the present invention.

[0023] Figure 3 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0025] Figure 1 This is a flowchart illustrating the method for dynamically determining the dosage of reagents in a heavy-medium high-density sedimentation tank provided by the present invention.

[0026] like Figure 1 As shown in this embodiment, a method for dynamically determining the dosage of reagents in a heavy-medium high-density sedimentation tank is provided. This method is applied to a continuous-flow, fully mixed heavy-medium high-density sedimentation tank, and is particularly suitable for coal washing wastewater treatment in coal preparation plants. The applicability is based on the premise that the water quality parameters inside the sedimentation tank are uniformly distributed, and that both PAC hydrolysis and PAM adsorption bridging reactions are irreversible. The reaction rate is positively correlated with the reagent concentration and water quality parameters. The floc particles carried by the returned sludge have a synergistic enhancing effect on coagulation and flocculation, without additional reagent consumption. Reagent consumption only comes from reactions corresponding to the SS and turbidity of the raw water, and the flocs removed by sedimentation do not release reagents again. The method mainly includes the following steps: 101. Collect the influent turbidity, effluent turbidity, influent flow rate, and sludge return ratio of the heavy media high-density sedimentation tank.

[0027] Specifically, the above parameters are all collected in real time by the corresponding online monitoring instruments. The collection frequency can be adjusted according to the degree of fluctuation in the operating conditions. Under normal operating conditions, the data is collected once per minute, and the collection frequency can be increased when the fluctuations are severe.

[0028] 102. Based on the dynamic change of the difference between influent turbidity and effluent turbidity, generate the turbidity treatment capacity change rate; based on the real-time fluctuation of influent flow rate, generate the influent flow rate change rate; based on the deviation between turbidity treatment capacity and standard operating condition treatment load, generate the treatment load change rate; based on the real-time change of sludge return ratio, generate the return ratio change rate.

[0029] Specifically, the difference between the influent turbidity and the effluent turbidity at the current moment is calculated, and then the rate of change of this difference over time is calculated to obtain the turbidity treatment capacity change rate. This parameter reflects the fluctuation of turbidity treatment demand per unit time.

[0030] Calculate the difference between the current influent flow rate and the influent flow rate under standard operating conditions, and then calculate the rate of change of this difference over time to obtain the influent flow rate change rate. This parameter reflects the degree of impact fluctuation in the influent flow rate.

[0031] Calculate the difference between the turbidity treatment load at the current moment and the turbidity treatment load under standard operating conditions, and then calculate the rate of change of this difference over time to obtain the rate of change of treatment load. This parameter reflects the dynamic change of the system's treatment pressure.

[0032] Calculate the difference between the sludge return ratio at the current moment and the sludge return ratio under standard operating conditions, and then calculate the rate of change of this difference over time to obtain the rate of change of the return ratio. This parameter reflects the dynamic change of the synergistic coagulation effect of the returned sludge.

[0033] 103. Based on the initial PAC dosage concentration, the PAC dosage concentration at the current moment is dynamically determined by combining the turbidity treatment rate change rate, influent flow rate change rate, treatment load change rate, reflux ratio change rate, and environmental correction coefficient. The environmental correction coefficient is determined by the pH correction coefficient, temperature correction coefficient, conductivity correction coefficient, and safety correction coefficient.

[0034] The PAC dosage concentration is as shown in formula (1): in, This indicates the concentration of PAC added at time t, in mg / L. The initial PAC concentration under standard operating conditions is determined by laboratory coagulation tests, and the unit is mg / L. This represents the weighting coefficient of the dynamic variable for PAC influent turbidity treatment, which is an empirical value. This represents the rate of change in turbidity treatment volume at time t, in NTU / h. This represents the weighting coefficient of the dynamic variable of PAC inflow rate, which is an empirical value. This represents the rate of change of influent flow rate at time t, compared to the standard operating flow rate, in cubic meters per second (m³). 3 / h 2 ; This represents the weighting coefficient for the PAC turbidity treatment load variable, which is an empirical value. This represents the rate of change of turbidity load at time t, compared to the standard operating condition load. This represents the weighting coefficient for the PAC sludge return ratio variable, which is an empirical value. This represents the rate of change of the reflux ratio at time t, compared to the reflux ratio under standard operating conditions. This represents the pH correction factor for PAC addition at time t; This represents the correction factor for the PAC injection temperature at time t; This represents the conductivity correction factor for PAC at time t; This represents the safety correction factor for PAC addition.

[0035] The calculation logic is as follows: taking the initial PAC concentration under standard operating conditions as a benchmark, the four types of dynamic change rates are multiplied by their corresponding weighting coefficients and then summed to obtain the basic ratio for dynamic adjustment. Then, the ratio is multiplied by four types of environmental correction coefficients: pH, temperature, conductivity, and safety, to obtain the dynamic change rate of the PAC concentration at the current moment. After integration, the real-time PAC concentration can be obtained, allowing the PAC concentration to be adjusted synchronously with the real-time fluctuations of water quality, water volume, and reflux ratio. At the same time, the influence of environmental factors on the coagulation effect is corrected, ensuring that the dosage is accurately matched with the real-time treatment requirements, and avoiding waste of reagents or substandard effluent quality.

[0036] 104. Output the corresponding PAC dosage control command according to the PAC dosage concentration.

[0037] Based on the real-time influent flow rate and PAC dosage concentration, the actual PAC dosage per unit time is calculated. This value is then sent as a control command to the on-site PLC or dosing control system to adjust the operating frequency of the dosing pump or the opening of the dosing valve, thereby achieving real-time dynamic adjustment of the PAC dosage. The calculation results are then converted into executable control actions to complete the closed-loop control of dynamic dosing.

[0038] 105. Real-time acquisition of the change rate of PAC dosage concentration and the change rate of magnetite sand dosage concentration. Based on the initial PAM dosage concentration, combined with the change rate of turbidity treatment volume, change rate of influent flow rate, change rate of PAC dosage concentration, change rate of magnetite sand dosage concentration, and PAM environmental correction coefficient, dynamically determine the current PAM dosage concentration.

[0039] Specifically, the rate of change in PAC dosage concentration For example, in formula (2): (2) The higher the PAC dosage, the more complete the destabilization of the colloids, the higher the PAM adsorption bridging efficiency, and the faster the reaction rate. These two factors show a positive correlation, with the index 0.8 obtained from empirical data fitting (adjusted based on different water quality pilot-scale conditions). The magnetite sand dosage ranges from 50 to 200 mg / L, and its concentration change rate is the rate of change of the magnetite sand dosage at the current moment. As a heavy medium assisted in flocculation, changes in the dosage of magnetite sand directly affect the floc settling effect.

[0040] The consumption of PAM is related to the adsorption and binding of destabilized particles, and the amount of PAM added needs to be matched with that of PAC. The typical ratio of PAC:PAM is 50:1 to 100:1. The formula for calculating the concentration of PAM is as follows (3): (3) in, This indicates the concentration of PAM added at time t, in mg / L. The initial PAM concentration under standard operating conditions is determined by laboratory testing and is expressed in mg / L. This represents the weighting coefficient of the dynamic variable for PAM influent turbidity treatment, which is an empirical value. This represents the weighting coefficient of the dynamic variable of PAM inflow rate, which is an empirical value. This represents the weighting coefficient for the PAC dosage concentration variable, which is an empirical value. This represents the rate of change in PAC concentration at time t; This represents the weighting coefficient of the PAM remediation variable, which is an empirical value. This represents the rate of change in the concentration of magnetite sand added at time t; This represents the pH correction factor for PAM addition at time t; This represents the correction factor for PAM application temperature at time t; This represents the conductivity correction factor for PAM at time t; This represents the safety correction factor added to PAM.

[0041] By synchronously and dynamically matching the PAM dosage with the PAC dosage and operating condition fluctuations, the coagulation aid effect is ensured to be compatible with the coagulation process, avoiding insufficient coagulation aid or waste of reagents caused by the imbalance of the two ratios.

[0042] Furthermore, based on the above embodiments, the determination of the pH correction coefficient in this embodiment includes: determining the pH correction coefficient based on the deviation of the real-time pH value from the preset optimal pH range and the pH change rate; determining the temperature correction coefficient includes: determining the temperature correction coefficient based on the deviation of the real-time temperature from the preset optimal temperature range and the temperature change rate, and triggering external temperature control measures when the real-time temperature exceeds the preset safe temperature threshold; determining the conductivity correction coefficient includes: calculating the conductivity correction coefficient based on the deviation of the real-time conductivity from the preset optimal conductivity range and the conductivity change rate, and triggering electrolyte addition or dilution measures when the real-time conductivity exceeds the preset safe conductivity threshold; determining the safety correction coefficient includes: determining the safety correction coefficient by weighted summation based on the fluctuation range of the influent flow rate, the fluctuation range of the influent water quality, the error of the turbidity detection instrument, and the purity deviation of the PAC reagent.

[0043] Specifically, the formula for calculating the pH correction factor is (4): (4) in, The pH sensitivity coefficient ranges from 0.18 to 0.25, with a typical value of 0.22. This represents the real-time pH value of the reaction tank at time t. This indicates the optimal pH value for PAC reaction efficiency, with a typical value of 7.8; This represents the compensation dosage coefficient when the pH value changes abruptly, with a value ranging from 0.02 to 0.04, and a typical value of 0.03. This represents the absolute value of the pH change rate, typically 0.075 for stable conditions and 0.25 for fluctuating conditions.

[0044] The optimal pH range for PAC hydrolysis is 7.5-8.5. The further the pH deviates from this range, the lower the PAC hydrolysis efficiency, requiring an increased dosage to compensate. Furthermore, rapid pH fluctuations necessitate additional compensation to offset the effects of sudden changes. The correction ranges and engineering implications for different pH ranges are shown in Table 1 below. Table 1

[0045] when When the pH value is greater than 2.5, the corresponding pH value is either below 5.5 or above 10.5. At this point, the cost of simply increasing the PAC dosage is higher than the cost of adding a neutralizing agent. The system should first add sulfuric acid or sodium hydroxide to bring the pH back to the range of 6.5-9.5, and then use a pH correction coefficient to adjust the dosage to accurately compensate for the impact of pH fluctuations on PAC hydrolysis efficiency, while also taking into account the economic efficiency of operation.

[0046] The temperature correction factor is adopted as a piecewise quadratic function, and the calculation formula is as follows (5): (5) in, The temperature of the reaction tank at time t is expressed in °C, and the operating limit range is 5-40 °C. This represents the nonlinear correction coefficient for temperature deviation, with a value ranging from 0.008 to 0.012, and a typical value of 0.010. This represents the linear correction factor for temperature deviation, with a value ranging from 0.02 to 0.03, and a typical value of 0.025. This indicates the lower limit of the optimal temperature; a typical value is 20°C. This indicates the upper limit of the optimal temperature, with a typical value of 25°C.

[0047] The mechanism by which temperature affects PAC efficiency is as follows: Optimal temperature range: 20~25℃ (typical value for coal preparation wastewater), at which temperature the PAC hydrolysis rate is the fastest, the amount of polynuclear hydroxyl complexes generated is the largest, and the coagulation efficiency is the highest; Low temperature suppression (T<20℃= The hydrolysis reaction rate decreases, the floc formation is slow and the particles are small, the sedimentation efficiency is reduced, and the dosage needs to be increased. High temperature suppression (T>25℃= ): It accelerates the decomposition of PAC hydrolysis products and the reduction of effective ingredients. At the same time, the Brownian motion of colloidal particles is intensified, making them difficult to aggregate, so the dosage needs to be increased.

[0048] When the real-time temperature is below 5℃, PAC hydrolysis almost stops, and the effect of chemical compensation is extremely poor. The system automatically triggers heating measures. The temperature correction coefficient is activated again after the temperature rises above 10℃. When the real-time temperature is above 40℃, the effective components of PAC decompose rapidly. The system triggers cooling measures. The temperature correction coefficient is activated again after the temperature drops below 35℃ to offset the impact of temperature fluctuations on the PAC reaction rate and effective components. Under extreme temperatures, external control ensures that the coagulation reaction can proceed normally.

[0049] The conductivity correction coefficient is adopted in the form of a piecewise quadratic function, and the calculation formula is as follows (6): (6) in, This represents the real-time conductivity at time t, in μS / cm. This represents a nonlinear correction factor for conductivity deviation, typically with a value of 2 × 10⁻⁷. This represents a linear correction factor for conductivity deviation, typically with a value of 1.2 × 10⁻⁴. This indicates the lower limit of the optimal conductivity, typically 500 μS / cm; This indicates the upper limit of the optimal conductivity, with a typical value of 2000 μS / cm.

[0050] Electrical conductivity (EC) reflects the ion concentration in water. The optimal conductivity range is 500-2000 μS / cm. Too low a conductivity weakens the charge neutralization effect, while too high a conductivity inhibits PAC hydrolysis; in both cases, the dosage needs to be increased to compensate. Under normal operating conditions, the correction factor ranges from 1.15 to 1.30 when the conductivity is below 500 μS / cm, 1.00 when the conductivity is between 500 and 2000 μS / cm, and 1.20 to 1.50 when the conductivity is above 2000 μS / cm.

[0051] When the real-time conductivity is below 300 μS / cm, the severe ion deficiency leads to the failure of charge neutralization. The system automatically adds electrolytes to adjust the conductivity to above 500 μS / cm. When the real-time conductivity is above 5000 μS / cm, the high-salt environment severely inhibits hydrolysis, and chemical compensation is ineffective. The system triggers dilution measures. After the conductivity drops below 5000 μS / cm, the conductivity correction coefficient is activated to compensate for the impact of differences in ion concentration in the water on the charge neutralization capacity of PAC. Under extreme conductivity conditions, external control is used to ensure the coagulation effect.

[0052] The formula for calculating the safety correction factor is as shown in formula (7): (7) in, This represents the weighting coefficient for water volume fluctuations, typically with a value of 0.3. Indicates the rate of change of water volume; This represents the water quality fluctuation weighting coefficient, typically with a value of 0.3. Indicates the range of turbidity fluctuation; This represents the error weighting coefficient of the online turbidity monitoring instrument, with a typical value of 0.2. This indicates the error value of the turbidity meter; This represents the weighting coefficient for drug purity, typically with a value of 0.2. This indicates the deviation value of PAC reagent purity.

[0053] The safety correction factor is based on 1. The four influencing factors are multiplied by their corresponding weights and then summed to reserve a reasonable safety margin for the dosage, so as to deal with uncertainties such as operating condition fluctuations, instrument errors, and reagent quality fluctuations, and avoid the effluent water quality exceeding the standard under extreme operating conditions.

[0054] Furthermore, based on the above embodiments, this embodiment determines the PAM environmental correction coefficient by: using an exponential function, based on the deviation of real-time pH from the optimal pH center of PAM, the deviation of real-time temperature from the optimal temperature range of PAM, and the deviation of real-time conductivity from the optimal conductivity range of PAM, and in combination with the molecular chain response hysteresis time, respectively determining the pH correction coefficient, temperature correction coefficient, and conductivity correction coefficient of PAM as the PAM environmental correction coefficient; after determining the PAM environmental correction coefficient, it also includes: controlling the PAM dosage concentration not to exceed the preset upper limit of the PAC dosage concentration.

[0055] Specifically, the coagulation aid effect of PAM is determined by the extension state of the molecular chain. When environmental parameters change, there is a time lag in the conformational adjustment of PAM molecular chains. Therefore, all corrections introduce the lag time parameter.

[0056] pH correction factor As in formula (8): (8) in, The value represents the molecular chain extension coefficient; the recommended value for anionic PAM is 0.13, for cationic PAM is 0.15, and for nonionic PAM is 0.10. This indicates the optimal pH value center; the recommended value for anionic PAM is 7.8, for cation PAM is 7.0, and for nonionic PAM is 7.5. This indicates the pH change response lag time, with a range of 3 to 5 minutes and a typical value of 4 minutes. The recommended value for PAM is 0.02 for anions, 0.03 for cations, and 0.01 for nonions. This represents a constant correction term, ensuring that the correction factor is 1 when the pH deviation is 0. The correction factor is 0.13 for anionic PAM, 0.15 for cations, and 0.10 for nonionic PAM.

[0057] The relationship between PAM (polyacrylamide) dosage concentration and pH value essentially stems from the fact that pH affects molecular chain conformation, adsorption bridging efficiency, and necessitates concentration adjustments for compensation. The relationship equation must satisfy the following: based on the optimal pH range, the greater the deviation, the larger the concentration correction factor. This is applicable to pH fluctuations >0.1 pH units / minute, such as during raw coal switching or abnormal dosing, requiring the addition of a pH change rate correction term to match the time lag in molecular chain extension. =3~5 minutes, usually 4 minutes.

[0058] Temperature correction factor As shown in formula (9): (9) in, This represents the temperature deviation nonlinearity correction factor, with a value range of 0.09-0.13 and a typical value of 0.11. This indicates the lower limit of the optimal temperature for PAM. This indicates the upper limit of the optimal temperature for PAM; This indicates the lag time in response to temperature changes, with a range of 3 to 5 minutes and a typical value of 4 minutes. This represents the correction factor for the rate of temperature change, with a value ranging from 0.01 to 0.03, and a typical value of 0.02. This represents a constant correction term, ensuring that the correction factor is 1 at the optimal temperature.

[0059] Temperature indirectly alters the coagulation efficiency of PAM by affecting its molecular chain conformation, viscosity characteristics, and diffusion rate. At the optimal temperature, the molecular chains are fully extended and diffused uniformly, resulting in the best coagulation effect. Low temperatures cause the molecular chains to shrink and the viscosity to increase, while high temperatures trigger the thermal degradation of the molecular chains. Both of these require compensation by increasing the dosage concentration through a dynamic correction coefficient.

[0060] The dynamic equations must satisfy: correction coefficients ≥1, the larger the coefficient is as far as the temperature deviates from the optimal temperature; includes time lag term and temperature change rate term; suitable for anionic PAM in coal preparation plants, molecular weight 8 million to 12 million, mainstream selection, parameters can be directly embedded into PLC / DCS control system.

[0061] Sudden temperature change, such as At that time, the PAM molecular chains cannot respond in time, resulting in a delayed attenuation of the coagulation-aiding effect. The correction coefficient is increased in advance to avoid slow settling of flocs in the effluent due to delayed molecular chain response.

[0062] Electrical conductivity, by reflecting the ion concentration in water, affects the molecular chain dissolution and diffusion characteristics, colloidal double layer state, and floc stability of PAM. Within the optimal EC range, with moderate ion concentration, PAM molecular chains are fully extended and diffused uniformly, resulting in the best coagulation efficiency. Low EC leads to thickening of the colloidal double layer, making flocs difficult to aggregate. High EC compresses the double layer, making flocs easily break. Both require compensation by increasing the dosage concentration through a dynamic correction coefficient.

[0063] Anionic PAM is less sensitive to EC than pH and temperature, but the effect is significant at low EC (<800 μS / cm) or high EC (>5000 μS / cm), requiring careful correction.

[0064] The conductivity correction factor adopts a piecewise exponential function form, and the calculation formula is as follows (10): (10) in, This represents the nonlinearity correction factor for conductivity, with a value ranging from 0.08 to 0.12, and a typical value of 0.10. This indicates the lower limit of the optimal conductivity of PAM, typically 500 μS / cm; This indicates the upper limit of the optimal conductivity of PAM, with a typical value of 2000 μS / cm; This indicates the hysteresis time of the response to changes in conductivity, with a range of 3 to 5 minutes and a typical value of 4 minutes. This represents the correction factor for the rate of change of conductivity, with a value range of 0.00008 to 0.00012, and a typical value of 0.0001. This represents a constant correction term. The correction factor is 1 to ensure optimal conductivity, with a typical value of 0.10.

[0065] When EC changes abruptly, PAM molecular chains cannot respond to changes in ion concentration in a timely manner, resulting in a delayed decay of the coagulation aid effect. By increasing the correction coefficient in advance, the settling velocity of flocs in the effluent can be avoided due to the delayed response of the molecular chains.

[0066] PAM safety factor K sBased on environmental factor correction, it is a comprehensive redundancy coefficient set for four core uncertainties: water quality fluctuation, water volume impact, turbidity change, and reagent purity deviation. It is used to ensure the stable coagulation effect of PAM under extreme working conditions, with effluent turbidity ≤5NTU, while avoiding cost waste and secondary pollution caused by excessive addition.

[0067] The formula for calculating the PAM safety correction factor is as follows (11): (11) in, This represents the weighting coefficient for water volume fluctuations, typically with a value of 0.3. This represents the water quality fluctuation weighting coefficient, typically with a value of 0.3. This represents the error weighting coefficient of the online turbidity monitoring instrument, with a typical value of 0.2. This represents the weighting coefficient for drug purity, typically with a value of 0.2. This indicates the purity deviation value of the PAM reagent.

[0068] After determining the PAM environmental correction factor, a dosage ratio constraint is set simultaneously, as shown in formula (12): (12) Excessive PAM addition can lead to the dissolution of flocs, which can reduce the sedimentation effect. Therefore, a limit is set on the proportion, and the PAM addition concentration shall not exceed 0.02 times the PAC addition concentration. This accurately compensates for the influence of environmental factors on the molecular chain state of PAM. At the same time, the proportion constraint avoids the side effects of excessive PAM addition and ensures the stability of the coagulant aid effect.

[0069] Furthermore, based on the above embodiments, after dynamically determining the PAM dosage concentration at the current moment, this embodiment further includes: determining whether the PAC dosage concentration and PAM dosage concentration exceed the corresponding preset hard constraint boundary, and determining the total cost rate in real time. The total cost rate includes at least one of the PAC cost rate, PAM cost rate, magnetite sand cost rate, and effluent quality exceeding penalty cost rate; when the PAC dosage concentration and / or PAM dosage concentration exceed the corresponding preset hard constraint boundary, an alarm is triggered and the dosage adjustment range is limited; when the total cost rate exceeds the preset cost upper limit, a cost optimization mode is activated to limit the increase of dosage concentration while ensuring that the water quality meets the standards.

[0070] Specifically, the hard constraint boundary is the process limit that must be met, as shown in formula (13): (13) When the concentration of PAC and / or PAM exceeds the corresponding hard constraint boundary, the system triggers an alarm and limits the adjustment range of the dosage, prohibiting the dosage from being adjusted further in the direction of exceeding the limit.

[0071] The total cost rate is the total operating cost per unit time, including the PAC cost rate, PAM cost rate, magnetite sand cost rate, and the penalty cost rate for exceeding effluent quality standards. The dynamic calculation formulas for each component cost rate are as follows: PAC cost rate dynamic equation (14): (14) in, This is the unit price for PAC, typically 2.8 yuan / kg; The unit conversion factor is used to convert mg / L to m 3 Convert the product of / h to kg / h.

[0072] The physical meaning is that the change in PAC cost rate is contributed by two parts: the dynamic adjustment of the dosage concentration. and traffic fluctuations .

[0073] PAM cost rate dynamic equation (15): (15) in, The price is for PAM, typically 18 yuan / kg.

[0074] Note: PAM dosage concentration is in ppm, which is equivalent to mg / m³. 3 The unit of influent flow rate is m. 3 Multiplying by / h gives mg / h, then by The coefficient is converted to kg / h to match the unit price.

[0075] Dynamic equation for magnetite cost rate (16): (16) in, The price is for magnetite sand, typically 0.4 yuan / kg.

[0076] The total cumulative cost is the integral of the cost rate over time and is used for long-term cost assessment. The complete total cumulative cost also includes power costs, sludge treatment costs, and other operating costs. The complete formula is shown in (17): (17) In real-time engineering calculations, auxiliary cost items can be selected to be included based on project requirements. The core dosing optimization mainly uses reagent costs and over-limit penalty costs as the main accounting items. The simplified formula is as follows, which can reduce computational complexity and is suitable for real-time control scenarios: (18) Under operating conditions where the flow rate change rate does not exceed 5% per minute, the calculation error of this simplified formula is less than 3%, which meets the engineering accuracy requirements.

[0077] Simultaneously, a soft cost constraint is set, as shown in formula (19): (19) in, The maximum allowable cost rate is set according to the project budget and can be dynamically adjusted based on fluctuations in operating conditions. When the total cost rate exceeds the preset cost limit, the cost optimization mode is activated. Under the premise of ensuring water quality compliance, the increase in dosage concentration is restricted. Priority is given to reducing costs by optimizing the dosage ratio and adjusting parameter matching. Water quality must not be sacrificed to reduce costs. Hard constraints ensure process safety and water quality compliance, while soft cost constraints control operating costs on the basis of compliance, balancing water quality and economy.

[0078] Furthermore, based on the above embodiments, this embodiment, after determining the penalty cost rate for exceeding the effluent quality standards, also includes: when the effluent turbidity exceeds the first turbidity threshold or the effluent suspended solids concentration exceeds the first concentration threshold, controlling the rate of change of the penalty cost rate for exceeding the effluent quality standards to increase dynamically in proportion to the rate of change of the extent of exceeding the standards.

[0079] Specifically, the formula for calculating the penalty cost rate for exceeding the standard is as follows (20): (20) in, This represents the penalty coefficient for exceeding turbidity standards, typically set at 50 yuan / (NTU·h), which can be adjusted based on wastewater treatment fees and penalties for exceeding standards.

[0080] When the effluent turbidity exceeds 5 NTU or the effluent suspended solids concentration exceeds 15 mg / L, the rate of change of the penalty cost rate for exceeding the effluent quality standards increases dynamically and proportionally to the rate of change of the degree of exceedance. The greater the degree of exceedance and the faster the increase in the exceedance value, the faster the penalty cost rate increases. When the effluent quality returns to within the acceptable range, the penalty cost rate returns to zero. The purpose of setting the penalty cost rate is to force the control logic to prioritize ensuring that the effluent quality meets the standards through a cost penalty mechanism, preventing the system from allowing water quality to exceed the standards in order to reduce reagent costs. It transforms the water quality compliance constraint into a quantifiable cost indicator, strengthens the water quality priority principle in the cost optimization process, and ensures that cost optimization does not come at the expense of exceeding water quality standards.

[0081] Furthermore, based on the above embodiments, this embodiment initiates a cost optimization mode, including: comparing the current total cost rate with the historical minimum cost rate; if the current total cost rate is lower than the historical minimum cost rate and the ratio of the two is less than a preset threshold, triggering a breakthrough reward; determining the breakthrough reward according to the logarithmic function of the ratio of the current total cost rate to the historical minimum cost rate, and updating the historical minimum cost rate based on the result of the logarithmic function.

[0082] Specifically, after the cost optimization mode is activated, the current total cost rate is continuously compared with the historical minimum cost rate under the corresponding operating condition. If the current total cost rate is lower than the historical minimum cost rate, and the ratio of the two is less than a preset threshold, a breakthrough reward is triggered. The calculation formula for the breakthrough reward is as follows (21): (twenty one) in, This indicates the breakthrough reward coefficient, typically 200. This represents the minimum cost rate of memory under the current operating conditions, expressed in yuan / hour. This indicates a breakout threshold, typically 0.95, meaning a breakout is triggered when the current cost is lower than 95% of the historical minimum cost.

[0083] The greater the cost reduction, the higher the breakthrough reward value. After the breakthrough reward is triggered, the historical minimum cost rate under the corresponding working condition is updated with the current total cost rate as the benchmark for optimization. The reward mechanism incentivizes the system to continuously explore lower cost operation schemes, continuously optimize the dosage ratio, and achieve a continuous reduction in operating costs.

[0084] Furthermore, based on the above embodiments, this implementation updates the historical minimum cost rate based on the result of the logarithmic function, including: when the breakthrough reward calculated by the logarithmic function exceeds the preset reward threshold, storing the current working condition feature vector, the current total cost rate, and the current addition amount combination as a new optimal record in the memory bank, and updating the historical minimum cost rate with the current total cost rate; clustering and merging the historical records in the memory bank according to the similarity between the working condition feature vectors, and eliminating historical records that have not been matched for a long time.

[0085] Specifically, when the breakthrough reward calculated by the logarithmic function exceeds the preset reward threshold, it indicates that the current solution is significantly optimized compared to the historical best solution. At this time, the current working condition feature vector, the current total cost rate, and the current addition amount are stored as a new optimal record in the memory bank, and the historical minimum cost rate of the corresponding working condition is updated with the current total cost rate. The storage structure formula of the memory bank is as follows (22): (twenty two) in, The feature vector representing the i-th type of working condition is the standardized data. This represents the minimum cost rate under the i-th type of operating condition, in yuan / h; This indicates the optimal dosage adjustment action corresponding to the minimum cost.

[0086] The data memory is regularly maintained and optimized, including: daily clustering optimization during the low-load period in the early morning; merging data conditions with a similarity higher than 0.9 into one class based on the similarity between their feature vectors, retaining the lowest-cost record in that class to reduce redundant data; automatically discarding historical records that have not been matched or retrieved for more than 180 days to prevent outdated and invalid data from affecting matching efficiency; and sorting the data memory by the frequency of occurrence of data conditions, prioritizing the matching of frequently occurring conditions to improve decision-making response speed.

[0087] By establishing an iterative optimal experience base, we continuously accumulate low-cost operating solutions, while ensuring the simplicity and timeliness of the memory base through regular maintenance.

[0088] Furthermore, based on the above embodiments, this embodiment further includes the following steps before updating the historical minimum cost rate based on the logarithmic function result: constructing a memory bank, which stores historical operating condition feature vectors, historical minimum cost rates, and historical optimal dosage combinations; performing cosine similarity matching between the current operating condition feature vector and the historical operating conditions in the memory bank; when a matching operating condition with a similarity higher than a preset threshold is found, calling the corresponding historical optimal dosage combination as a reference; when no matching operating condition with a similarity higher than the preset threshold is found, determining the optimal dosage combination through a reinforcement learning model with the goal of minimizing the total cost, and storing the current operating condition features, the corresponding minimum cost rate, and the optimal dosage combination into the memory bank after stable operation.

[0089] Specifically, when new operating condition data is received, the cosine similarity between the current operating condition feature vector and all historical operating conditions in the memory is calculated to determine the degree of matching. The cosine similarity calculation formula is as follows (23): (twenty three) in, The feature vector representing the current new working condition. It is represented as the feature vector of the i-th type of historical working condition in the memory bank.

[0090] When a historical operating condition with a similarity higher than 0.85 is matched, the historical optimal dosage combination corresponding to that operating condition is directly called as the reference scheme for the current operation. During the operation, the cost and water quality performance of the scheme are continuously verified. If the cost of the currently adjusted scheme is lower and the water quality meets the standards, the memory record of the corresponding operating condition is updated. When no historical operating condition with a similarity higher than 0.85 is matched, it indicates that it is a new operating condition. At this time, the reinforcement learning model is used to explore and determine the optimal dosage combination with the goal of minimizing the total cost. The state space formula of the reinforcement learning model is as follows (24): (twenty four) The state space includes real-time water quality, process parameters, dosage, and cost rate, providing a comprehensive understanding of the current operating conditions.

[0091] The action space formula for a reinforcement learning model is shown in (25): (25) in, This indicates the dosage adjustment amount, with an adjustment range of -5 to +5 mg / L and a step size of 0.5 mg / L; This indicates the adjustment amount of PAM dosage, with an adjustment range of -0.2 to +0.2 ppm and a step size of 0.05 ppm; This indicates the adjustment amount of magnetite sand dosage, with an adjustment range of -10 to +10 mg / L and a step size of 5 mg / L.

[0092] After the reinforcement learning-output solution under the new operating condition has been running stably for 30 minutes, the current operating condition characteristics, the corresponding minimum cost rate, and the optimal dosage combination are stored in the memory bank for subsequent use under similar operating conditions. Mature and optimal solutions can be directly called upon for common operating conditions, resulting in fast response times. For unfamiliar operating conditions, reinforcement learning is used to autonomously explore and optimize, balancing operational efficiency and continuous optimization capabilities.

[0093] Furthermore, based on the above embodiments, the reinforcement learning model in this embodiment adopts a reward function, which includes at least one of a basic reward, a target reward, and a breakthrough reward; wherein, the basic reward is negatively correlated with the current total cost rate, and the target reward increases with the improvement of water quality when the effluent water quality is better than the preset standard.

[0094] Specifically, the general formula for the reward function used in reinforcement learning models is as follows (26): (26) The reward function consists of a basic reward, a compliance reward, a breakthrough reward, and a penalty term, which guides the model to iteratively optimize towards achieving water quality standards and minimizing costs.

[0095] Basic reward, i.e., cost minimization orientation, such as (27): (27) in, This represents the base reward coefficient, with an experience value of 100. Cost rate. The lower the base reward, the higher the incentive to reduce costs.

[0096] Water quality compliance rewards, i.e., hard constraints and guarantees, such as (28): (28) in, =50 is the basic reward for meeting the target. =30 is the turbidity optimization reward coefficient. =10 is the SS optimization reward coefficient.

[0097] The better the water quality is than the standard, the higher the reward, encouraging the model to optimize the effluent water quality on the basis of meeting the standard and avoid fluctuations at the critical level of meeting the standard.

[0098] Minimum cost breakthrough reward, i.e. memory reinforcement guidance, as mentioned above (21), when the cost of the found operating solution is lower than the historical minimum cost and the breakthrough condition is met, an additional breakthrough reward is obtained, which incentivizes the model to continue exploring better solutions.

[0099] Penalties, i.e., penalties for violating constraints, such as (29): -5)+ -15)+ (29) in, =200 indicates the penalty coefficient for exceeding the turbidity standard; =100 indicates the penalty coefficient for exceeding the SS limit; =300 indicates the penalty coefficient for exceeding the dosage limit; This is an indicator function; it takes the value 1 when the amount added exceeds the hard constraint boundary, and 0 otherwise.

[0100] The penalty value is much higher than the regular reward, ensuring that the model prioritizes hard constraints and will not exceed process or water quality limits in order to reduce costs. Through a multi-dimensional reward and penalty mechanism, the optimization direction and constraint boundary of reinforcement learning are clearly defined, ensuring that the solution output by the model meets both water quality and process requirements and achieves optimal cost.

[0101] Compared with traditional static dosing methods, the method of this invention has the following core advantages: Both the cost model and the dosage model employ dynamic calculations to adapt to real-time fluctuations in water quality and flow, offering higher calculation accuracy than static cost models. They integrate three types of constraints: process boundaries, water quality requirements, and cost ceilings, ensuring that the final dosing decision achieves both compliance and economic efficiency. Through operating condition matching and memory update mechanisms, the system can directly access historical best practices, eliminating the need for repeated exploration and improving decision-making efficiency. The memory bank stores the correspondence between operating conditions, costs, and dosages, providing a traceable basis for every optimal decision.

[0102] From an engineering application perspective, based on a processing flow rate of 5000m³ / h... 3 Based on typical operating conditions such as raw water SS=100mg / L and turbidity=50NTU, the AI ​​control method using this approach can reduce overall costs by 15%~20%, resulting in annual cost savings of approximately 150,000~250,000 yuan. Simultaneously, it reduces the subjectivity and lag of manual adjustments, lowering the standard deviation of effluent water quality by over 30%. Ultimately, it achieves full automation of water quality prediction, dosing decisions, cost optimization, and memory learning, reducing reliance on operator experience.

[0103] Figure 2 This is a schematic diagram of the system for dynamically determining the dosage of reagents in a heavy medium high-density sedimentation tank provided by the present invention.

[0104] Based on the same general inventive concept, such as Figure 2 As shown, this invention also protects a system for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank, comprising: The data acquisition module 201 is used to collect the influent turbidity, effluent turbidity, influent flow rate, and sludge return ratio of the heavy media high-density sedimentation tank. The generation module 202 is used to generate a turbidity treatment capacity change rate based on the dynamic change of the difference between the influent turbidity and the effluent turbidity, generate an influent flow rate change rate based on the real-time fluctuation of the influent flow rate, generate a treatment load change rate based on the deviation between the turbidity treatment capacity and the standard operating condition treatment load, and generate a sludge return ratio change rate based on the real-time change of the sludge return ratio. The first determining module 203 is used to dynamically determine the current PAC dosage concentration based on the initial PAC dosage concentration, combined with the turbidity treatment rate change rate, influent flow rate change rate, treatment load change rate, reflux ratio change rate, and environmental correction coefficient; wherein, the environmental correction coefficient is jointly determined by the pH correction coefficient, temperature correction coefficient, conductivity correction coefficient, and safety correction coefficient. Output module 204 is used to output a corresponding PAC dosage control command according to the PAC dosage concentration; The second determining module 205 is used to acquire the change rate of PAC dosage concentration and the change rate of magnetite sand dosage concentration in real time; based on the initial PAM dosage concentration, combined with the change rate of turbidity treatment volume, the change rate of influent flow rate, the change rate of PAC dosage concentration, the change rate of magnetite sand dosage concentration and the PAM environmental correction coefficient, the PAM dosage concentration at the current moment is dynamically determined.

[0105] Figure 3 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 3As shown, the electronic device may include: a processor 310, a communication interface 320, a memory 330, and a communication bus 340. The processor 310, communication interface 320, and memory 330 communicate with each other via the communication bus 340. The processor 310 can call logical instructions in the memory 330 to execute a method for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank. This method includes: collecting the influent turbidity, effluent turbidity, influent flow rate, and sludge return ratio of the heavy media high-density sedimentation tank; generating a turbidity treatment rate change based on the dynamic change in the difference between the influent and effluent turbidity; generating an influent flow rate change based on the real-time fluctuation of the influent flow rate; generating a treatment load change rate based on the deviation between the turbidity treatment rate and the standard operating condition treatment load; generating a return ratio change rate based on the real-time change of the sludge return ratio; and using the initial PAC dosage concentration as a benchmark, combined with the turbidity treatment rate change, influent flow rate change, and... The system dynamically determines the PAC dosage concentration at the current moment by considering the load change rate, reflux ratio change rate, and environmental correction coefficient. The environmental correction coefficient is jointly determined by pH correction coefficient, temperature correction coefficient, conductivity correction coefficient, and safety correction coefficient. Based on the PAC dosage concentration, a corresponding PAC dosage control command is output. The system also acquires the change rates of PAC and magnetite sand dosage in real time. Using the initial PAM dosage concentration as a benchmark, and combining the turbidity treatment rate change, influent flow rate change, PAC dosage concentration change, magnetite sand dosage concentration change, and PAM environmental correction coefficient, the system dynamically determines the PAM dosage concentration at the current moment.

[0106] Furthermore, the logical instructions in the aforementioned memory 330 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0107] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the dynamic determination method for the dosage of reagents in a heavy media high-density sedimentation tank provided by the above methods, including: collecting the influent turbidity, effluent turbidity, influent flow rate, and sludge return ratio of the heavy media high-density sedimentation tank; generating a turbidity treatment capacity change rate based on the dynamic change of the difference between the influent turbidity and the effluent turbidity; generating an influent flow rate change rate based on the real-time fluctuation of the influent flow rate; generating a treatment load change rate based on the deviation between the turbidity treatment capacity and the standard operating condition treatment load; and generating a return ratio change rate based on the real-time change of the sludge return ratio; with P Based on the initial AC dosage concentration, the PAC dosage concentration at the current moment is dynamically determined by combining the turbidity treatment rate change rate, influent flow rate change rate, treatment load change rate, reflux ratio change rate, and environmental correction coefficient. The environmental correction coefficient is jointly determined by pH correction coefficient, temperature correction coefficient, conductivity correction coefficient, and safety correction coefficient. According to the PAC dosage concentration, a corresponding PAC dosage control command is output. The change rates of PAC dosage concentration and magnetite sand dosage concentration are acquired in real time. Based on the initial PAM dosage concentration, and combining the turbidity treatment rate change rate, influent flow rate change rate, PAC dosage concentration change rate, magnetite sand dosage concentration change rate, and PAM environmental correction coefficient, the PAM dosage concentration at the current moment is dynamically determined.

[0108] Furthermore, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon. When executed by a processor, this computer program implements a method for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank, as provided by the methods described above. This method includes: collecting the influent turbidity, effluent turbidity, influent flow rate, and sludge return ratio of the heavy media high-density sedimentation tank; generating a turbidity treatment rate change rate based on the dynamic change in the difference between the influent and effluent turbidity; generating an influent flow rate change rate based on real-time fluctuations in the influent flow rate; generating a treatment load change rate based on the deviation between the turbidity treatment rate and the standard operating condition treatment load; and generating a return ratio change rate based on real-time changes in the sludge return ratio; using the initial PAC dosage concentration as a benchmark, combined with... The PAC dosage concentration at the current moment is dynamically determined by the turbidity treatment rate change rate, influent flow rate change rate, treatment load change rate, reflux ratio change rate, and environmental correction coefficient. The environmental correction coefficient is jointly determined by pH correction coefficient, temperature correction coefficient, conductivity correction coefficient, and safety correction coefficient. A corresponding PAC dosage control command is output according to the PAC dosage concentration. The change rates of PAC dosage concentration and magnetite sand dosage concentration are acquired in real time. Using the initial PAM dosage concentration as a benchmark, and combining the turbidity treatment rate change rate, influent flow rate change rate, PAC dosage concentration change rate, magnetite sand dosage concentration change rate, and PAM environmental correction coefficient, the PAM dosage concentration at the current moment is dynamically determined.

[0109] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0110] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0111] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank, characterized in that, include: Collect the influent turbidity, effluent turbidity, influent flow rate, and sludge return ratio of the heavy media high-density sedimentation tank; Based on the dynamic change of the difference between the influent turbidity and the effluent turbidity, a turbidity treatment capacity change rate is generated; based on the real-time fluctuation of the influent flow rate, an influent flow rate change rate is generated; based on the deviation between the turbidity treatment capacity and the standard operating condition treatment load, a treatment load change rate is generated; and based on the real-time change of the sludge return ratio, a return ratio change rate is generated. Based on the initial PAC dosage concentration, and combined with the turbidity treatment rate change rate, influent flow rate change rate, treatment load change rate, reflux ratio change rate, and environmental correction coefficient, the current PAC dosage concentration is dynamically determined; wherein, the environmental correction coefficient is jointly determined by the pH correction coefficient, temperature correction coefficient, conductivity correction coefficient, and safety correction coefficient. According to the PAC dosage concentration, output the corresponding PAC dosage control command; The PAM dosage concentration change rate and the magnetite sand dosage change rate are acquired in real time. Based on the initial PAM dosage concentration, the PAM dosage concentration at the current moment is dynamically determined by combining the turbidity treatment rate change, influent flow rate change, PAC dosage concentration change, magnetite sand dosage concentration change, and PAM environmental correction coefficient.

2. The method for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank according to claim 1, characterized in that, Determining the pH correction factor includes: determining the pH correction factor based on the deviation of the real-time pH value from the preset optimal pH range and the pH change rate; Determining the temperature correction coefficient includes: determining the temperature correction coefficient based on the deviation of the real-time temperature from the preset optimal temperature range and the rate of temperature change, and triggering external temperature control measures when the real-time temperature exceeds the preset safe temperature threshold. Determining the conductivity correction coefficient includes: calculating the conductivity correction coefficient based on the deviation between the real-time conductivity and the preset optimal conductivity range and the conductivity change rate, and triggering electrolyte addition or dilution measures when the real-time conductivity exceeds the preset safe conductivity threshold; The safety correction factor is determined by weighted summation based on the fluctuation range of influent flow rate, the fluctuation range of influent water quality, the error of turbidity detection instrument, and the purity deviation of PAC reagent.

3. The method for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank according to claim 1, characterized in that, Determining the PAM environmental correction factor includes: Using an exponential function, based on the deviation of real-time pH from the optimal pH center of PAM, the deviation of real-time temperature from the optimal temperature range of PAM, and the deviation of real-time conductivity from the optimal conductivity range of PAM, and combined with the molecular chain response hysteresis time, the pH correction coefficient, temperature correction coefficient, and conductivity correction coefficient of PAM are determined respectively, and used as the PAM environmental correction coefficient. After determining the PAM environmental correction coefficient, the method further includes controlling the PAM dosage concentration to not exceed the preset upper limit of the PAC dosage concentration.

4. The method for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank according to claim 1, characterized in that, After dynamically determining the current PAM dosage concentration, the method further includes: Determine whether the PAC and PAM dosage concentrations exceed the corresponding preset hard constraint boundaries, and determine the total cost rate in real time. The total cost rate includes at least one of the PAC cost rate, PAM cost rate, magnetite sand cost rate, and effluent quality exceeding penalty cost rate. When the concentration of PAC and / or the concentration of PAM exceed the corresponding preset hard constraint boundary, an alarm is triggered and the dosage adjustment range is limited. When the total cost rate exceeds the preset cost limit, the cost optimization mode is activated, and the dosage concentration is limited to increase while ensuring that the water quality meets the standards.

5. The method for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank according to claim 4, characterized in that, After determining the penalty cost rate for exceeding the effluent quality standards, the following is also included: When the turbidity of the effluent exceeds the first turbidity threshold or the concentration of suspended solids in the effluent exceeds the first concentration threshold, the rate of change of the penalty cost rate for exceeding the effluent quality standard increases dynamically in proportion to the rate of change of the extent of exceeding the standard.

6. The method for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank according to claim 4, characterized in that, The startup cost optimization mode includes: Compare the current total cost rate with the historical minimum cost rate. If the current total cost rate is lower than the historical minimum cost rate and the ratio of the two is less than a preset threshold, trigger a breakthrough reward. The breakthrough reward is determined according to the logarithmic function of the ratio of the current total cost rate to the historical minimum cost rate, and the historical minimum cost rate is updated based on the result of the logarithmic function.

7. The method for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank according to claim 6, characterized in that, Updating the historical minimum cost rate based on the result of the logarithmic function includes: When the breakthrough reward calculated by the logarithmic function exceeds the preset reward threshold, the current working condition feature vector, the current total cost rate and the current addition amount combination are stored as a new optimal record in the memory bank, and the historical minimum cost rate is updated with the current total cost rate. Based on the similarity between the feature vectors of operating conditions, the historical records in the memory are clustered and merged, and historical records that have not been matched for a long time are eliminated.

8. The method for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank according to claim 7, characterized in that, Before updating the historical minimum cost rate based on the result of the logarithmic function, the method further includes: Construct a memory library, which stores historical operating condition feature vectors, historical minimum cost rates, and historical optimal combination of addition amounts; Perform cosine similarity matching between the current working condition feature vector and the historical working conditions in the memory; When a working condition with a similarity higher than a preset threshold is matched, the corresponding historical best dosage combination is used as a reference. When no matching operation condition with similarity higher than the preset threshold is found, the optimal dosage combination is determined by the reinforcement learning model with the goal of minimizing the total cost. After stable operation, the current operation condition characteristics, the corresponding minimum cost rate, and the optimal dosage combination are stored in the memory bank.

9. The method for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank according to claim 8, characterized in that, The reinforcement learning model employs a reward function, which includes at least one of a basic reward, a target achievement reward, and a breakthrough reward. The basic reward is negatively correlated with the current total cost rate, and the compliance reward increases as the water quality improves when the effluent quality is better than the preset standard.

10. A system for dynamically determining the dosage of reagents in a heavy media high-density sedimentation tank, characterized in that, include: The data acquisition module is used to collect the influent turbidity, effluent turbidity, influent flow rate, and sludge return ratio of the heavy media high-density sedimentation tank. The generation module is used to generate a turbidity treatment capacity change rate based on the dynamic change of the difference between the influent turbidity and the effluent turbidity, to generate an influent flow rate change rate based on the real-time fluctuation of the influent flow rate, to generate a treatment load change rate based on the deviation between the turbidity treatment capacity and the standard operating condition treatment load, and to generate a sludge return ratio change rate based on the real-time change of the sludge return ratio. The first determining module is used to dynamically determine the current PAC dosage concentration based on the initial PAC dosage concentration, combined with the turbidity treatment rate change rate, influent flow rate change rate, treatment load change rate, reflux ratio change rate, and environmental correction coefficient; wherein, the environmental correction coefficient is jointly determined by the pH correction coefficient, temperature correction coefficient, conductivity correction coefficient, and safety correction coefficient. The output module is used to output the corresponding PAC dosage control command according to the PAC dosage concentration; The second determining module is used to acquire the change rate of PAC dosage concentration and the change rate of magnetite sand dosage concentration in real time; based on the initial PAM dosage concentration, combined with the change rate of turbidity treatment volume, the change rate of influent flow rate, the change rate of PAC dosage concentration, the change rate of magnetite sand dosage concentration and the PAM environmental correction coefficient, the PAM dosage concentration at the current moment is dynamically determined.

Images