Method, system and equipment for optimizing dosage of chemicals in semiconductor factory affair system

Abstract

The invention discloses a method, system and equipment for optimizing the dosage of chemicals in a semiconductor factory affair system, and the method comprises the following steps: obtaining historical data which comprises the water inlet flow, the water inlet PH value, the dosage of chemicals and the water outlet PH value; using a time sequence model to predict the water inlet flow and the water inlet PH value at a plurality of moments in the future; calculating a predicted effluent PH value through a polynomial nonlinear regression model based on the predicted effluent flow, the predicted effluent PH value and the chemical use amount; solving and optimizing the chemical dosage under constraint conditions by using a nonlinear programming solving method and taking minimization of the chemical cost as a target; and outputting the optimized chemical dosage. According to the method, multivariable time series data can be efficiently, accurately and systematically processed, a complex nonlinear relation is modeled, and cost-optimal chemical dosage optimization can be realized under actual constraints.

Description

Technical Field

[0001] This invention belongs to the field of semiconductor manufacturing technology, and in particular relates to a method, system and equipment for optimizing the amount of chemicals used in a semiconductor plant system. Background Technology

[0002] Semiconductor plant systems are the "infrastructure and support systems" that underpin semiconductor chip manufacturing. They are auxiliary but crucial supporting systems, distinct from chip design and wafer fabrication (core processes such as photolithography and etching). Their core function is to provide semiconductor production lines with stable, precise, and stringent-standard environmental conditions, resource supply, and environmental protection measures, ensuring the stability of the chip manufacturing process and product yield.

[0003] In the semiconductor manufacturing industry, the pure water preparation and wastewater treatment processes in plant systems frequently require the use of various chemicals (such as HCl and NaOH) to adjust water quality parameters (such as pH) and ensure that the effluent meets stringent process standards. Currently, the control of chemical dosing in such systems largely relies on experience or predictions based on simple statistical models, which has significant limitations. While existing technologies employ time-series forecasting or regression analysis for water quality or dosage prediction, they often use only a single model and fail to effectively integrate the nonlinear relationships and interactions between multiple variables, resulting in limited prediction accuracy. Furthermore, the lack of an automated optimization mechanism that combines prediction results with actual engineering constraints (such as water quality specifications and cost targets) makes it difficult to achieve precise optimization of chemical dosage and cost control. In addition, existing solutions are mostly implemented with customized code, resulting in low model iteration and maintenance efficiency, insufficient engineering stability, and an inability to meet the high real-time and multi-constraint decision-making requirements of the semiconductor manufacturing environment. Therefore, there is an urgent need for a chemical dosage optimization method that can efficiently, accurately, and systematically process multivariate time-series data, model complex nonlinear relationships, and achieve cost-optimal results under actual constraints. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a method, system, equipment and medium for optimizing chemical usage in a semiconductor plant system, which can efficiently, accurately and systematically process multivariate time series data, model complex nonlinear relationships, and achieve cost-optimized chemical usage under actual constraints.

[0005] The first aspect of this invention discloses a method for optimizing chemical usage in a semiconductor plant system, comprising the following steps: Acquire historical data, including influent flow rate, influent pH value, chemical dosage, and effluent pH value; The influent flow rate and influent pH value are predicted at multiple future time points using time series models, wherein the time series models are built based on historical data; Based on the predicted influent flow rate, predicted influent pH value, and chemical dosage, the predicted effluent pH value is calculated using a polynomial nonlinear regression model, wherein the polynomial nonlinear regression model includes linear terms and cross terms for chemical dosage, influent flow rate, and influent pH value. The nonlinear programming solution method is used to optimize the amount of chemicals used under constraints, with the goal of minimizing the cost of chemicals. These constraints include the amount of chemicals used being greater than zero and the predicted pH value of the effluent not exceeding the preset specification value. Output the optimized chemical dosage.

[0006] The above method uses a time series model, including at least one of an ARIMA model, a Prophet model, an LSTM model, or a Seq2seq model, for multi-time-point prediction of influent flow rate and influent pH value. The mathematical expression of the time series model is as follows:

[0007] Among them, X t This represents the value of the variable at the current moment, i.e., the influent flow rate or the influent pH value, a. j X represents the regression coefficient. t-j ε represents the value of the variable at a historical moment. t denoted by , where p represents the error term and p represents the model order.

[0008] The mathematical expression of the polynomial nonlinear regression model described above is:

[0009] Where y represents the predicted effluent pH value, a1, a2, a3, and a4 represent regression coefficients, x1 and x2 represent chemical dosage factors, m represents the predicted influent flow rate, n represents the predicted influent pH value, and the cross term represents the interaction term between chemical dosage, influent flow rate, and influent pH value.

[0010] In the above method, the cross term includes at least one of the following: the product of chemical dosage and influent flow rate, the product of chemical dosage and influent pH value, and the product of influent flow rate and influent pH value.

[0011] The mathematical expression of the above-described nonlinear programming solution method includes the objective function and constraints: Objective function:

[0012] Where Z represents the total cost, p i Let x represent the unit price of the i-th chemical. i Indicates the amount of the i-th chemical used; Constraints:

[0013] Among them, spec indicates the maximum specified value of the effluent pH value.

[0014] In the above method, the multiple future times include times t1, t2, and t3, and the time series model outputs the predicted values ​​of influent flow rate and influent pH value at these times.

[0015] The above method also includes deploying the chemical usage optimization method in the semiconductor plant system as an application programming interface (API) to receive real-time influent flow rate and influent pH data and return optimized chemical usage; the API is automatically built and published through a low-code platform.

[0016] The second aspect of this invention discloses a chemical usage optimization system in a semiconductor plant system, comprising a data acquisition module, a time series prediction module, a polynomial prediction module, a programming solution module, and an output module; The data acquisition module is used to acquire historical data, including influent flow rate, influent pH value, chemical dosage, and effluent pH value. The time series prediction module is used to predict the influent flow rate and influent pH value at multiple future times using a time series model, wherein the time series model is built based on historical data; The polynomial prediction module is used to calculate the predicted effluent pH value based on the predicted influent flow rate, the predicted influent pH value, and the chemical dosage using a polynomial nonlinear regression model, wherein the polynomial nonlinear regression model includes linear terms and cross terms for the chemical dosage, influent flow rate, and influent pH value. The planning and solving module is used to solve for the optimal chemical usage under constraints, with the goal of minimizing chemical costs, using a nonlinear programming solution method. The constraints include chemical usage being greater than zero and the predicted effluent pH value not exceeding a preset specification value. The output module is used to output the optimized chemical dosage.

[0017] A third aspect of the present invention discloses an electronic device, comprising: a memory and a processor, wherein the processor and the memory are connected; The memory is used to store programs; The processor invokes a program stored in the memory to execute the method provided by the first aspect embodiment and / or any possible embodiment in combination with the first aspect embodiment.

[0018] The fourth aspect of the present invention discloses a computer-readable storage medium having a computer program stored thereon, the computer program being executed by a computer to perform the methods provided by the embodiments of the first aspect and / or any possible embodiments in combination with the embodiments of the first aspect.

[0019] Compared with existing technologies, this invention has the following advantages: By introducing an organic combination of multivariate time series prediction and nonlinear regression modeling, it effectively solves the key problem of chemical dosing control in the water treatment process of semiconductor plant systems. This invention utilizes a time series model to predict influent flow rate and influent pH value, accurately capturing the dynamic changing trends of water quality parameters and overcoming the lag and inaccuracy of traditional empirical operations or simple statistical models. Furthermore, through a polynomial nonlinear regression model (including linear and cross terms of chemical dosage, influent flow rate, and influent pH value), it explicitly characterizes the complex nonlinear interactions between multiple variables (such as the synergistic effect of chemical dosing and water flow rate, and the nonlinear response to pH value changes), significantly improving the prediction accuracy of effluent pH value and avoiding the deficiency of a single model in fitting complex relationships. Based on this, a nonlinear programming solution method is adopted, with the goal of minimizing chemical costs and the constraint that the effluent pH value does not exceed the preset specification value, achieving cost-optimal decision-making under strict process standards. This automatically combines the prediction results with actual engineering constraints (water quality specifications, cost targets), solving the problem of the lack of optimization mechanisms in traditional methods. Furthermore, the proposed model framework is structured and iterative (e.g., the separation and integration of time series and regression models), reducing reliance on customized code, improving model operation and maintenance efficiency and engineering stability, and enabling it to adapt to the high real-time requirements of semiconductor manufacturing environments (e.g., real-time adjustment of chemical dosage) and multi-constraint decision-making needs. Ultimately, through systematic data-driven optimization, this method significantly reduces chemical usage while ensuring that the effluent quality meets standards, achieving precise cost control, and simultaneously enhancing the system's adaptability.

[0020] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0021] Figure 1 This is a flowchart of the method in Embodiment 1 of the present invention.

[0022] Figure 2 This is a system module diagram of Embodiment 2 of the present invention. Detailed Implementation

[0023] Example 1 like Figure 1 As shown, a method for optimizing chemical usage in a semiconductor plant system includes the following steps: Step 1: Obtain historical data, including influent flow rate, influent pH value, chemical dosage, and effluent pH value; During implementation, historical data is collected from the plant's system database or sensors, including influent flow rate, influent pH value, chemical dosage (such as the amount of HCl and NaOH added), and effluent pH value. This data is typically stored in time-series format, for example, recorded every minute or hour. In practice, the data is extracted via data interfaces or ETL tools and then cleaned and preprocessed, such as handling missing or outlier values. For example, in a pure water system scenario, influent flow rate, influent pH value, HCl dosage, NaOH dosage, and effluent pH value data for the past three months are obtained.

[0024] Step 2: Use a time series model to predict the influent flow rate and influent pH value at multiple future times, wherein the time series model is built based on historical data; In implementation, a time series model, such as the ARIMA model, is built based on historical data. This model utilizes autoregressive and moving average components to capture trends and seasonality in the data. During model training, historical influent flow rate and pH value data are used as input, and the model parameters are fitted using maximum likelihood estimation or least squares method. During prediction, the model outputs predicted influent flow rate and pH value for multiple future time points (e.g., 1 hour, 2 hours, and 3 hours). By accurately predicting future influent conditions, it provides input for subsequent calculations. For example, in a pure water system scenario, the ARIMA model is used to predict the influent flow rate and pH value at future times t1, t2, and t3, and these predicted values ​​are used in the next calculation.

[0025] Step 3: Based on the predicted influent flow rate, predicted influent pH value, and chemical dosage, calculate the predicted effluent pH value using a polynomial nonlinear regression model, wherein the polynomial nonlinear regression model includes linear and cross terms for chemical dosage, influent flow rate, and influent pH value. In implementation, the multinomial nonlinear regression model is built on historical data. The model expression includes linear and interaction terms for chemical dosage, influent flow rate, and influent pH value. During model training, historical data is used to fit the regression coefficients, for example, through least squares optimization or gradient descent. When calculating the predicted effluent pH value, the predicted influent flow rate, predicted influent pH value, and assumed chemical dosage are substituted into the model to obtain the predicted effluent pH value. By simulating the nonlinear relationship between the effluent pH value and the input variables, the prediction accuracy is improved. For example, in a pure water system scenario, the model expression is: Where y is the predicted effluent pH, x1 is the HCl dosage, x2 is the NaOH dosage, m is the predicted influent flow rate, n is the predicted influent pH, and a1 to a4 are regression coefficients. Based on the predicted m and n values, and the assumed x1 and x2, the y value is calculated.

[0026] Step 4: Using a nonlinear programming solution method, with the goal of minimizing chemical costs, optimize the amount of chemicals used under constraints, including chemical usage being greater than zero and the predicted effluent pH value not exceeding the preset specification value. In implementation, the nonlinear programming solution uses minimizing chemical costs as the objective function. For example, the objective function is min Z = p1x1 + p2x2, where p1 and p2 are the unit prices of HCl and NaOH, and x1 and x2 are the amounts of chemicals used. Constraints include chemical usage being greater than zero and the predicted effluent pH value not exceeding a preset specification value (e.g., effluent pH value ≤ 7.5). During the solution process, iterative algorithms such as interior-point methods or sequential quadratic programming are used to adjust the values ​​of x1 and x2 until the minimum cost solution satisfying the constraints is found. By optimizing the amount of chemicals added, operating costs are reduced while ensuring effluent quality. For example, in a pure water system scenario, given the predicted influent flow rate and influent pH value, the solution aims to minimize the cost of HCl and NaOH usage while ensuring the predicted effluent pH value is ≤ 7.5.

[0027] Step 5: Output the optimized chemical dosage.

[0028] During implementation, the optimized chemical dosage obtained from the solution is output to the control system or display interface for automatic control of the chemical dosing pump or to provide operational guidance. The output can be in the form of digital values, messages, or commands. This enables real-time optimized control and improves operational efficiency. For example, in a pure water system scenario, the optimized HCl and NaOH dosage values ​​are output to the plant execution system to automatically adjust the dosing rate.

[0029] In this embodiment, the time series model includes at least one of the following: ARIMA model, Prophet model, LSTM model, or Seq2seq model, used for multi-time-point prediction of influent flow rate and influent pH value. The mathematical expression of the time series model is as follows:

[0030] Among them, X t This represents the value of the variable at the current moment, i.e., the influent flow rate or the influent pH value, a. j X represents the regression coefficient. t-j ε represents the value of the variable at a historical moment. t denoted by , where p represents the error term and p represents the model order.

[0031] In implementation, choose a model type suitable for the data: ARIMA models are suitable for stationary time series, while differencing handles non-stationary data; Prophet models handle seasonality and holiday effects; LSTM or Seq2seq models handle long-term dependencies. Model training uses historical data, and performance is optimized through hyperparameter tuning. During prediction, the model outputs influent flow rate and influent pH value for multiple future time points. The technical benefits include improved prediction accuracy and robustness, adapting to different data characteristics. For example, in a pure water system scenario, using an ARIMA model to predict influent flow rate (p=3), based on data from the past 24 hours, predicts the value for the next 3 hours.

[0032] In this embodiment, the mathematical expression of the polynomial nonlinear regression model is:

[0033] Where y represents the predicted effluent pH value, a1, a2, a3, and a4 represent regression coefficients, x1 and x2 represent chemical dosage factors, m represents the predicted influent flow rate, n represents the predicted influent pH value, and the cross term represents the interaction term between chemical dosage, influent flow rate, and influent pH value.

[0034] In implementation, historical data is used for model training, and coefficient values ​​are determined through regression analysis. During calculation, predicted values ​​and assumed values ​​are substituted to obtain y. The technical effect is to capture the interactions between variables and improve model fit. For example, in a pure water system scenario, the expression is specifically y = 0.5x1 - 0.3x2 + 0.2m + 0.1n + 0.05x1m - 0.02x1n + 0.03x2m + 0.01x2n + 0.001mn, with coefficients obtained by fitting historical data.

[0035] In this embodiment, the cross term includes at least one of the following: the product of chemical dosage and influent flow rate, the product of chemical dosage and influent pH value, and the product of influent flow rate and influent pH value.

[0036] In implementation, these product terms are explicitly added as features to the multinomial regression model. During model training, these product terms participate in coefficient fitting. This reduces prediction errors by better simulating complex nonlinear relationships. For example, in a pure water system scenario, the interaction term, such as x1×m, represents the interaction between HCl dosage and influent flow rate in affecting the effluent pH value.

[0037] In this embodiment, the mathematical expression of the nonlinear programming solution method includes the objective function and constraints: Objective function:

[0038] Where Z represents the total cost, p iLet x represent the unit price of the i-th chemical. i Indicates the amount of the i-th chemical used; Constraints:

[0039] Among them, spec indicates the maximum specified value of the effluent pH value.

[0040] In implementation, optimization algorithms such as fmincon in MATLAB or scipy.optimize in Python are used to iteratively find the optimal solution. For example, in a pure water system scenario, the objective function is min Z = 0.1x1 + 0.2x2 (assuming the unit price of HCl is 0.1 yuan / liter and the unit price of NaOH is 0.2 yuan / liter), with constraints x1>0, x2>0, and y ≤ 7.5.

[0041] In this embodiment, the multiple future times include times t1, t2, and t3, and the time series model outputs the predicted values ​​of influent flow rate and influent pH value at these times.

[0042] In practice, the time series model predicts the influent flow rate and pH value at specific moments, which are then used for subsequent calculations. For example, in a pure water system scenario, the values ​​of t1 (after 1 hour), t2 (after 2 hours), and t3 (after 3 hours) are predicted, and the chemical dosage at each moment is optimized accordingly.

[0043] In this embodiment, the method for optimizing chemical usage in the semiconductor plant system is also deployed as an application programming interface (API) to receive real-time influent flow rate and influent pH data and return optimized chemical usage. The API is automatically built and published through a low-code platform.

[0044] In implementation, a low-code platform (such as OutSystems or Mendix) is used to encapsulate the chemical usage optimization method for the entire semiconductor plant system into a RESTful API. The API receives real-time influent flow rate and pH value data and returns the optimized chemical usage. The platform automatically handles model calls, calculations, and deployments. For example, in a pure water system scenario, the API obtains real-time data via HTTP requests, calls the internal model to calculate, and returns the optimized usage. The plant system can automatically adjust chemical additions through API calls.

[0045] Example 2 like Figure 2 As shown, a chemical usage optimization system in a semiconductor plant includes a data acquisition module, a time series prediction module, a polynomial prediction module, a programming solution module, and an output module. The data acquisition module is used to acquire historical data, including influent flow rate, influent pH value, chemical dosage, and effluent pH value. The time series prediction module is used to predict the influent flow rate and influent pH value at multiple future times using a time series model, wherein the time series model is built based on historical data; The polynomial prediction module is used to calculate the predicted effluent pH value based on the predicted influent flow rate, the predicted influent pH value, and the chemical dosage using a polynomial nonlinear regression model, wherein the polynomial nonlinear regression model includes linear terms and cross terms for the chemical dosage, influent flow rate, and influent pH value. The planning and solving module is used to solve for the optimal chemical usage under constraints, with the goal of minimizing chemical costs, using a nonlinear programming solution method. The constraints include chemical usage being greater than zero and the predicted effluent pH value not exceeding a preset specification value. The output module is used to output the optimized chemical dosage.

[0046] The chemical usage optimization system in the semiconductor plant system provided in this embodiment has the same implementation principle and technical effect as the method embodiment in Embodiment 1. For the sake of brevity, any parts not mentioned in the system embodiment can be referred to the corresponding content in Embodiment 1.

[0047] Example 3 A computer-readable storage medium having a computer program stored thereon, the computer program being executed by a computer to perform the chemical usage optimization method in a semiconductor plant system described in Embodiment 1 above.

[0048] Example 4 An electronic device includes: a memory and a processor, wherein the processor and the memory are connected; The memory is used to store programs; The processor invokes a program stored in the memory to execute a chemical usage optimization method in a semiconductor plant system as described in Example 1.

[0049] It should be noted that the electronic device mentioned may be, but is not limited to, personal computers (PCs), tablet computers, mobile internet devices (MIDs), etc.

[0050] It should be noted that processors, memory, and other components that may be present in electronic devices are electrically connected to each other, directly or indirectly, to enable data transmission or interaction. For example, processors, memory, and other components may be electrically connected to each other via one or more communication buses or signal lines.

[0051] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0052] In the several embodiments provided in this application, it should be understood that the disclosed systems and methods can also be implemented in other ways. The system embodiments described above are merely illustrative. For example, the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram and / or flowchart, and combinations of blocks in block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0053] In addition, the functional modules in the various embodiments of this application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0054] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion 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, laptop, server, mobile phone, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. 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.

[0055] The above description is merely a preferred embodiment of the present invention and does not constitute any limitation on the present invention. Any simple modifications, alterations, or equivalent structural changes made to the above embodiments based on the technical essence of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for optimizing chemical usage in a semiconductor plant system, characterized in that, Includes the following steps: Acquire historical data, including influent flow rate, influent pH value, chemical dosage, and effluent pH value; The influent flow rate and influent pH value are predicted at multiple future time points using time series models, wherein the time series models are built based on historical data; Based on the predicted influent flow rate, predicted influent pH value, and chemical dosage, the predicted effluent pH value is calculated using a polynomial nonlinear regression model, wherein the polynomial nonlinear regression model includes linear terms and cross terms for chemical dosage, influent flow rate, and influent pH value. The nonlinear programming solution method is used to optimize the amount of chemicals used under constraints, with the goal of minimizing the cost of chemicals. These constraints include the amount of chemicals used being greater than zero and the predicted pH value of the effluent not exceeding the preset specification value. Output the optimized chemical dosage.

2. The method as described in claim 1, characterized in that, The time series model includes at least one of the following: ARIMA model, Prophet model, LSTM model, or Seq2seq model, used for multi-time-point prediction of influent flow rate and influent pH value, wherein the mathematical expression of the time series model is: Among them, X t This represents the value of the variable at the current moment, i.e., the influent flow rate or the influent pH value, a. j X represents the regression coefficient. t-j ε represents the value of a variable at a historical moment. t denoted by , where p represents the error term and p represents the model order.

3. The method as described in claim 1, characterized in that, The mathematical expression for the polynomial nonlinear regression model is: Where y represents the predicted effluent pH value, a1, a2, a3, and a4 represent regression coefficients, x1 and x2 represent chemical dosage factors, m represents the predicted influent flow rate, n represents the predicted influent pH value, and the cross term represents the interaction term between chemical dosage, influent flow rate, and influent pH value.

4. The method as described in claim 3, characterized in that, The cross term includes at least one of the following: the product of chemical dosage and influent flow rate, the product of chemical dosage and influent pH value, and the product of influent flow rate and influent pH value.

5. The method as described in claim 1, characterized in that, The mathematical expression of the nonlinear programming solution method includes the objective function and constraints: Objective function: Where Z represents the total cost, p i Let x represent the unit price of the i-th chemical. i Indicates the amount of the i-th chemical used; Constraints: Among them, spec indicates the maximum specified value of the effluent pH value.

6. The method as described in claim 1, characterized in that, The future multiple time points include t1, t2, and t3, and the time series model outputs the predicted influent flow rate and influent pH value at these time points.

7. The method as described in claim 1, characterized in that, It also includes deploying the chemical usage optimization method in the semiconductor plant system as an application programming interface (API) to receive real-time influent flow rate and influent pH data and return optimized chemical usage; the API is automatically built and published through a low-code platform.

8. A chemical usage optimization system in a semiconductor plant system, characterized in that, It includes a data acquisition module, a time series prediction module, a multinomial prediction module, a programming solution module, and an output module; The data acquisition module is used to acquire historical data, including influent flow rate, influent pH value, chemical dosage, and effluent pH value. The time series prediction module is used to predict the influent flow rate and influent pH value at multiple future times using a time series model, wherein the time series model is built based on historical data; The polynomial prediction module is used to calculate the predicted effluent pH value based on the predicted influent flow rate, the predicted influent pH value, and the chemical dosage using a polynomial nonlinear regression model, wherein the polynomial nonlinear regression model includes linear terms and cross terms for the chemical dosage, influent flow rate, and influent pH value. The planning and solving module is used to solve for the optimal chemical usage under constraints, with the goal of minimizing chemical costs, using a nonlinear programming solution method. The constraints include chemical usage being greater than zero and the predicted effluent pH value not exceeding a preset specification value. The output module is used to output the optimized chemical dosage.

9. An electronic device, characterized in that, include: A memory and a processor, wherein the processor and the memory are connected; The memory is used to store programs; The processor invokes a program stored in the memory to execute the method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, It stores a computer program, which is executed by a computer to perform the method as described in any one of claims 1-7.

Images