Intelligent carbon source dosing system and method with load sensing and sharing function

Abstract

This invention relates to wastewater treatment control technology, and discloses an intelligent carbon source dosing system and method for wastewater treatment with load sensing and allocation functions. The method includes: carbon source dosing, which involves adding carbon source to the anoxic tank and denitrification tank via a carbon source dosing pump unit; controlling the carbon source dosing frequency, which involves controlling the carbon source dosing frequency of the carbon source dosing pump unit via a carbon source dosing pump control cabinet; and feeding back the carbon source dosing amount to the carbon source dosing pump control cabinet; and determining the carbon source dosing amount, which is controlled by the intelligent carbon source dosing system through controlling the frequency of the carbon source dosing pump unit. The control of the carbon source dosing amount is achieved by controlling the carbon source dosing pump unit through the intelligent carbon source dosing system. This invention can detect the carbon source dosing amount in real time, and it offers good safety, low cost, and simple operation.

Description

Technical Field

[0001] This invention relates to wastewater treatment control technology, and more particularly to a wastewater intelligent carbon source dosing system and method with load sensing and allocation function. Background Technology

[0002] In wastewater treatment, the addition of carbon sources is to provide essential nutrients for microorganisms, promoting their absorption and transformation of nitrogen in wastewater, thereby achieving effective removal of total nitrogen and meeting nitrogen emission standards. Therefore, the rational addition of carbon sources is crucial.

[0003] Currently, most wastewater treatment plants control carbon source addition manually or through single-indicator feedback control. Manual control often involves staff adjusting the carbon source dosage based on indicators such as total nitrogen in the influent or effluent, combined with operational experience. While this method can maintain denitrification efficiency, it typically requires excessive addition, leading to increased operating costs. Single-indicator feedback control usually uses effluent total nitrogen as the indicator to control carbon source addition, resulting in significant lag in carbon source addition and an inability to comprehensively consider the impact of various water treatment indicators on microbial nitrogen removal efficiency, such as influent flow rate, temperature, ammonia nitrogen content, carbon source dosage, and water pH.

[0004] To address the drawbacks of manual control and single-indicator feedback control, existing technologies have been improved.

[0005] For example, Chinese patent CN202311335942.0 discloses an intelligent carbon source addition method. Based on the current influent flow rate, the ratio of carbon source addition to denitrification nitrogen removal, the nitrate nitrogen concentration difference, and the chemical oxygen demand equivalent of the carbon source, static correction coefficients, dynamic correction coefficients, and ratio values ​​are calculated. Finally, the above results are input into the carbon source calculation model to obtain the required carbon source addition. Although this method considers multiple water treatment indicators and reduces the carbon source addition to a certain extent, most of the coefficients required by the carbon source calculation model rely on manual experience calculations, making it difficult to adaptively handle changes in total nitrogen in the influent and the calculation is complex.

[0006] Existing technology 2: Chinese patent CN202310744254 discloses a method for determining the required carbon source dosage in real time based on the changing patterns and trends between input operating indicators and equipment control values, and then adjusting the carbon source dosage control strategy accordingly. However, it has poor adaptive capability and cannot learn the relationship between total nitrogen and nitrate nitrogen in real time. Summary of the Invention

[0007] This invention addresses the problem in existing technologies that struggle to adaptively and rapidly respond to changes in total nitrogen in influent and environmental conditions, thus hindering the achievement of optimal treatment results. It provides a wastewater intelligent carbon source dosing system and method with load sensing and allocation capabilities.

[0008] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0009] A wastewater intelligent carbon source dosing method with load sensing and allocation functions includes an anoxic tank, a carbon source dosing pump unit, an intelligent control cabinet for the carbon source dosing pump, and an intelligent carbon source dosing system; the method includes,

[0010] The carbon source is added to the anoxic tank through a carbon source dosing pump unit.

[0011] The carbon source dosing frequency is controlled by the carbon source dosing pump control cabinet, which controls the carbon source dosing pump unit's carbon source dosing frequency and transmits the carbon source dosing amount to the carbon source dosing pump control cabinet.

[0012] The amount of carbon source added is determined by controlling the frequency of the carbon source dosing pump unit through an intelligent carbon source dosing system.

[0013] As a preferred embodiment, the intelligent carbon source dosing system uses methods to control the carbon source dosing amount by controlling the frequency of the carbon source dosing pump unit, including:

[0014] The carbon source addition dataset was prepared by preprocessing the input raw time series dataset. The input raw time series dataset included the total nitrogen (TN) influent to the anoxic tank. in Concentration time series data, total nitrogen and total effluent TN out Concentration time series data, nitrate-nitrogen-anoxic tank NO3 concentration time series data, and anoxic tank influent flow rate in Time series data;

[0015] The intelligent carbon source dosing algorithm controls the carbon source dosing amount required for the corresponding anoxic pool load by using machine learning algorithms on the input raw time series dataset.

[0016] Preferably, the carbon source addition dataset is preprocessed using a clustering algorithm. The clustering algorithm processing includes:

[0017] Obtain influent and effluent parameters and data from the dosing pump over the past 24 hours, and calculate the average value X. ave ,

[0018]

[0019] Where, x i This refers to the i-th data point from the past.

[0020] Calculate the distance L between the current data and the average of the past 24 hours:

[0021]

[0022] Where x is the current data, X ave This is the average value of data from the past 24 hours.

[0023] Preferably, the carbon source dosage (Carbon) used to obtain the total nitrogen removal amount from the input raw time series dataset via machine learning algorithms includes:

[0024]

[0025] Among them, R TN ω represents the total amount of nitrogen to be removed, and ω is a model parameter.

[0026] For the amount of total nitrogen that needs to be removed, R TN The total nitrogen (TN) in the incoming water in (total) and total nitrate nitrogen (NO3) were obtained;

[0027] R TN =TN in (total)-NO3(total);

[0028] Among them, TN in (total) represents the total amount of nitrogen in the incoming water, and NO3(total) represents the total amount of nitrate nitrogen.

[0029] As a preferred option, for TN in (TOTEL) represents the total nitrogen content in the incoming water, measured by the incoming water TN. in and acquisition of incoming water flow;

[0030] TN in (total) = TN in *Flow in ;

[0031] Total nitrogen (NO3) is obtained by measuring nitrate nitrogen and influent flow rate.

[0032] NO3(total) = NO3 * Flow in ;

[0033] In this context, NO3 represents the amount of nitrate nitrogen.

[0034] As a preferred method, the amount of nitrate nitrogen (NO3) obtained is as follows:

[0035] A model was built based on the total nitrogen in the incoming water and the nitrate nitrogen in the anoxic tank. The model coefficients were obtained using the least squares method, and a real-time calculation method was used to ensure that the model reflects the relationship between total nitrogen and nitrate nitrogen at any given moment.

[0036] TN out =β0+β1TNin +β2NO3+ε

[0037] Where ε is the difference between the model's predicted value and the actual value; β0, β1 and β2 are the optimal parameters;

[0038] Fitting model coefficients using the least squares method To obtain the optimal parameters β0, β1, and β2:

[0039] By manually setting the target TN for the effluent index target Based on the current inflow index TN in By reverse calculation, the target value of nitrate nitrogen (NO3) in the anoxic pond can be determined.

[0040]

[0041] Among them, TN target Target values ​​for water discharge indicators are set manually.

[0042] As a preferred method, the optimal carbon source dosage will be calculated using gradient descent.

[0043] Calculate the loss function;

[0044]

[0045] Where λ is an adjustable parameter. The value of the penalty item.

[0046] Find the parameter ω that minimizes the loss function, and then differentiate the loss function to obtain the following formula:

[0047]

[0048] Where α is the learning rate, we find ω and λ when the loss function converges to its minimum, and confirm the weights and bias function; after obtaining the optimal parameters ω, we use ω to calculate the latest total nitrogen, nitrate nitrogen, and influent flow rate, as well as the latest total nitrogen removal rate R. TN Get the latest required carbon source dosage:

[0049]

[0050] Among them, R TN This is the latest total nitrogen removal rate.

[0051] As a preferred option, the carbon source addition amount is also adjusted, which is done through a dynamic learning strategy.

[0052] To address the aforementioned technical problems, this invention also provides a wastewater intelligent carbon source dosing system with load sensing and allocation functions, comprising an anoxic tank, a denitrification tank, a carbon source dosing pump unit, an intelligent control cabinet for the carbon source dosing pump, and an intelligent carbon source dosing system; it also includes,

[0053] The carbon source dosing module adds carbon source to the anoxic tank and denitrification tank through the carbon source dosing pump unit;

[0054] The carbon source dosing frequency control module controls the carbon source dosing frequency of the carbon source dosing pump unit through the carbon source dosing pump control cabinet; and feeds back the carbon source dosing amount to the carbon source dosing pump control cabinet.

[0055] The carbon source dosage determination module controls the carbon source dosage by controlling the frequency of the carbon source dosing pump unit through the intelligent carbon source dosing system.

[0056] The carbon source dosage control module controls the carbon source dosage by controlling the carbon source dosing pump unit through the intelligent carbon source dosing system.

[0057] This invention, by adopting the above technical solutions, has significant technical effects:

[0058] This invention can process changes in total nitrogen in wastewater in real time and is simple to calculate.

[0059] This invention adaptively and rapidly responds to changes in total nitrogen in the influent and changes in environmental conditions. After total nitrogen is removed in the anoxic tank, subsequent processes will not introduce new total nitrogen, so the relationship between total nitrogen in the effluent and nitrate nitrogen in the anoxic tank is stable. Attached Figure Description

[0060] Figure 1 This is a flowchart of the present invention.

[0061] Figure 2 This is a diagram illustrating the effect of nitrate and nitrogen control in this invention.

[0062] Figure 3 This is a diagram illustrating the effect of total nitrogen control in the effluent of this invention. Detailed Implementation

[0063] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments.

[0064] Example 1

[0065] A wastewater intelligent carbon source dosing method with load sensing and allocation functions includes an anoxic tank, a carbon source dosing pump unit, an intelligent control cabinet for the carbon source dosing pump, and an intelligent carbon source dosing system; the method includes,

[0066] The carbon source is added to the anoxic tank through a carbon source dosing pump unit.

[0067] The carbon source dosing frequency is controlled by the carbon source dosing pump control cabinet, which controls the carbon source dosing pump unit's carbon source dosing frequency and transmits the carbon source dosing amount to the carbon source dosing pump control cabinet.

[0068] The amount of carbon source added is determined by controlling the frequency of the carbon source dosing pump unit through an intelligent carbon source dosing system.

[0069] The intelligent carbon source dosing system uses methods to control the carbon source dosage by controlling the frequency of the carbon source dosing pump unit, including:

[0070] The carbon source addition dataset was prepared by preprocessing the input raw time series dataset. The input raw time series dataset included the total nitrogen (TN) influent to the anoxic tank. in Concentration time series data, total nitrogen and total effluent TN out Concentration time series data, nitrate-nitrogen-anoxic tank NO3 concentration time series data, and anoxic tank influent flow rate in Time series data;

[0071] The intelligent carbon source dosing algorithm controls the carbon source dosage required for the corresponding anoxic pool load by using machine learning algorithms on the input raw time series dataset.

[0072] The carbon source addition dataset was preprocessed using a clustering algorithm. The clustering algorithm processing included:

[0073] Obtain influent and effluent parameters and data from the dosing pump over the past 24 hours, and calculate the average value X. ave ,

[0074]

[0075] Where, x i This refers to the i-th data point from the past.

[0076] We need the average of the past 24 hours' data (sampled every minute), so the above formula requires 1440 data points for calculation. Next, we calculate the distance L between the current data and the average of the past 24 hours' data:

[0077]

[0078] Where x is the current data, X ave This is the average value of data from the past 24 hours.

[0079] We need a day's worth of data, with a sampling frequency of once per minute, totaling 1440 data points. We calculate the distance D between each of the 1440 data points and arrange them in ascending order. We then remove the 5% of data points with the largest distance D, i.e., 1440 * 5% = 72 data points, from the dataset. These data points are defined as outliers because they differ significantly from the average data of the past 24 hours. Including outliers in machine learning algorithms will cause the fitted model to be inconsistent with reality, so outliers need to be removed.

[0080] The carbon source dosage (Carbon) used to obtain the total nitrogen removal amount from the input raw time series dataset via machine learning algorithms includes:

[0081]

[0082] Among them, R TN ω represents the total amount of nitrogen to be removed, and ω is a model parameter.

[0083] For the amount of total nitrogen that needs to be removed, R TN The total nitrogen (TN) in the incoming water in (total) and total nitrate nitrogen (NO3) were obtained;

[0084] R TN =TN in (total)-NO3(total);

[0085] Among them, TN in (total) represents the total amount of nitrogen in the incoming water, and NO3(total) represents the total amount of nitrate nitrogen.

[0086] Since total nitrogen is the final effluent target, but the update time for total nitrogen values ​​is relatively long, system regulation based on total nitrogen values ​​would result in a lag. Therefore, the main invention of this patent lies in using nitrate nitrogen as the primary control factor. The rationale behind this method is that, due to process limitations, after total nitrogen is removed in the anoxic tank, subsequent processes will not introduce new total nitrogen, thus maintaining a stable relationship between effluent total nitrogen and anoxic tank nitrate nitrogen.

[0087] For TN in (total) represents the total nitrogen content in the incoming water, measured by the TN content of the incoming water. in and acquisition of incoming water flow;

[0088] TN in (total) = TN in *Flow in ;

[0089] Total nitrogen (NO3) is obtained by measuring nitrate nitrogen and influent flow rate.

[0090] NO3(total) = NO3 * Flow in ;

[0091] In this context, NO3 represents the amount of nitrate nitrogen.

[0092] The amount of nitrate nitrogen (NO3) obtained,

[0093] Because the total nitrogen in the effluent is affected by both the total nitrogen in the influent and the amount of carbon source added, but the amount of carbon source added is a variable that needs to be controlled, and according to the process logic, the nitrate nitrogen in the anoxic tank and the total nitrogen in the effluent are strongly correlated, we model the system based on the total nitrogen in the influent and the nitrate nitrogen in the anoxic tank, and derive the model coefficients using the least squares method.

[0094] A model was built based on the total nitrogen in the incoming water and the nitrate nitrogen in the anoxic tank. The model coefficients were obtained using the least squares method, and a real-time calculation method was used to ensure that the model reflects the relationship between total nitrogen and nitrate nitrogen at any given moment.

[0095] TN out =β0+β1TN in +β2NO3+ε

[0096] Where ε is the difference between the model's predicted value and the actual value; β0, β1 and β2 are the optimal parameters;

[0097] Fit the model coefficients using the least squares method. To obtain the optimal parameters β0, β1, and β2: by finding the gaps between all model predictions and actual values, find the β0, β1, and β2 corresponding to the time when they are minimized. These parameters are the optimal parameters for the model because the gap between the model predictions and actual values ​​is minimized under these parameters.

[0098] By manually setting the target TN for the effluent index target Based on the current inflow index TN in By reverse calculation, the target value of nitrate nitrogen (NO3) in the anoxic pond can be determined.

[0099]

[0100] Among them, TN target Target values ​​for effluent indicators are set manually. Because nitrate nitrogen levels in the anoxic tank provide more timely feedback, subsequent control will use the nitrate nitrogen target value calculated based on the total nitrogen target in the total effluent. The following method describes how to use machine learning to derive the relationship between nitrate nitrogen and carbon source dosage in the anoxic tank, and how to control the carbon source dosage based on this relationship.

[0101] The optimal carbon source dosage will be calculated using gradient descent.

[0102] Calculate the loss function;

[0103]

[0104] Where λ is an adjustable parameter. The value of the penalty item.

[0105] Find the parameter ω that minimizes the loss function, and then differentiate the loss function to obtain the following formula:

[0106]

[0107] Where α is the learning rate.

[0108] Find ω and λ when the loss function converges to its minimum, and confirm the weights and bias function; after obtaining the optimal parameter ω, use ω to calculate the latest total nitrogen, nitrate nitrogen, and influent flow rate, and the latest total nitrogen removal rate R. TN To obtain the latest required carbon source dosage (Carbon): After determining the optimal parameter ω, we use ω to calculate the total nitrogen removal amount R. TN Linked to the carbon source dosage (Carbon), the latest total nitrogen removal rate (R) is calculated using the latest influent total nitrogen, nitrate nitrogen, and influent flow rate. TN Calculate the latest required amount of carbon source to add;

[0109]

[0110] Among them, R TN This is the latest total nitrogen removal rate.

[0111] Example 2

[0112] Based on Example 1, this example also includes adjusting the input dataset and formulating an input dataset selection strategy through dynamic learning.

[0113] Obtain common values ​​for influent and effluent indicators and dosing pumps, and remove outliers with a distance L; calculate the distance L between the current data and the average of the past 24 hours' data.

[0114]

[0115] Where x is the current data, X ave This is the average value of data from the past 24 hours.

[0116] When the distance L data is determined to be an outlier, the previous data will be used to fill the gap.

[0117] Based on the new total nitrogen removal and historical carbon source addition, the new carbon source addition is calculated as shown in Table 1.

[0118] Table 1 Carbon Source Addition Amount

[0119] time Carbon source dosing pump flow rate 2024 / 08 / 2900:00:00 <![CDATA[x1]]> ... 2024 / 08 / 3023:58:00 <![CDATA[x 2879 <!-- 6 -->]]> 2024 / 08 / 3023:59:00 <![CDATA[x 2880 ]]>

[0120] x represents the flow rate of the carbon source dosing pump at that time point. Data is collected once per minute. At the latest data point, it is necessary to determine whether the data is an outlier.

[0121] Step 1: Calculate the average amount of carbon source added over the past day.

[0122]

[0123] Step 2: Calculate the distance D between each data point from the past day and the average value of the previous 24 hours, and take D as the distance. 2 ;

[0124]

[0125] Step 3: Calculate L 1440 ,...,L 2880 Sort by size from smallest to largest, and the top 5% will be removed as outlier data. Assumption: L 2880 and L 2879 All are within the highest 5%, so we will replace them with the previous data, as shown in Table 2.

[0126] Table 2 Data Preprocessing Table

[0127]

[0128]

[0129] Example 3

[0130] Based on the above embodiments, this embodiment is a wastewater intelligent carbon source dosing system with load sensing and allocation function, including an anoxic tank, a denitrification tank, a carbon source dosing pump unit, a carbon source dosing pump intelligent control cabinet, and an intelligent carbon source dosing system; it also includes,

[0131] The carbon source dosing module adds carbon source to the anoxic tank and denitrification tank through the carbon source dosing pump unit;

[0132] The carbon source dosing frequency control module controls the carbon source dosing frequency of the carbon source dosing pump unit through the carbon source dosing pump control cabinet; and feeds back the carbon source dosing amount to the carbon source dosing pump control cabinet.

[0133] The carbon source dosage determination module controls the carbon source dosage by controlling the frequency of the carbon source dosing pump unit through the intelligent carbon source dosing system.

[0134] The carbon source dosage control module controls the carbon source dosage by controlling the carbon source dosing pump unit through the intelligent carbon source dosing system.

Claims

1. A wastewater intelligent carbon source dosing method with load sensing and allocation function, including an anoxic tank, a carbon source dosing pump unit, an intelligent control cabinet for the carbon source dosing pump, and an intelligent carbon source dosing system; the method includes, The carbon source is added to the anoxic tank through a carbon source dosing pump unit. The carbon source dosing frequency is controlled by the carbon source dosing pump control cabinet, which controls the carbon source dosing pump unit's carbon source dosing frequency and transmits the carbon source dosing amount to the carbon source dosing pump control cabinet. The carbon source dosage is determined by controlling the frequency of the carbon source dosing pump unit through an intelligent carbon source dosing system. The intelligent carbon source dosing system uses methods to control the carbon source dosage by controlling the frequency of the carbon source dosing pump unit, including: The carbon source addition dataset is prepared by preprocessing the input raw time series dataset. The input raw time series dataset includes total nitrogen influent and concentration time series data, total nitrogen effluent concentration time series data, nitrate nitrogen anoxic tank concentration time series data, and anoxic tank influent flow rate time series data. The intelligent carbon source dosing algorithm controls the carbon source dosage required for the corresponding anoxic tank load using a machine learning algorithm on the input raw time series dataset; the carbon source dosage required for the total nitrogen removal when using a machine learning algorithm on the input raw time series dataset includes: ; Among them, R TN The amount of total nitrogen that needs to be removed. These are model parameters; For the amount of total nitrogen that needs to be removed, R TN It is obtained by measuring the total nitrogen and total nitrate nitrogen in the incoming water; ； Among them, TN in (total) represents the total amount of nitrogen in the incoming water, and NO3 (total) represents the total amount of nitrate nitrogen. For TN in (total) represents the total nitrogen content in the incoming water, measured by the TN content of the incoming water. in and acquisition of incoming water flow; ； Total nitrogen (NO3) is obtained by measuring nitrate nitrogen and influent flow rate. ； Wherein, NO3 represents the amount of nitrate nitrogen; Obtaining the amount of nitrate nitrogen (NO3): A model was built based on the total nitrogen in the incoming water and the nitrate nitrogen in the anoxic tank. The model coefficients were obtained using the least squares method, and a real-time calculation method was used to ensure that the model reflects the relationship between total nitrogen and nitrate nitrogen at any given moment. ； Where ℇ represents the difference between the model's predicted value and the actual value; β0, β1, and β2 are the optimal parameters; Fit the model coefficients using the least squares method. To obtain the optimal parameters β0, β1, and β2: By manually setting the target TN for the effluent index target Based on the current inflow index TN in By reverse calculation, the amount of nitrate and nitrogen (NO3) in the anoxic tank can be determined as follows: ； Among them, TN target Target values ​​for water discharge indicators are set manually.

2. The intelligent carbon source dosing method for wastewater with load sensing and allocation function according to claim 1, characterized in that, The carbon source addition dataset was preprocessed using a clustering algorithm. The clustering algorithm processing included: Obtain influent and effluent parameters and data from the dosing pump over the past 24 hours, and calculate the average value X. wave , ； Among them, X i This refers to the i-th data point from the past. Calculate the distance L between the current data and the average of the past 24 hours: ; Where X is the current data, X wave This is the average value of data from the past 24 hours.

3. The intelligent carbon source dosing method for wastewater with load sensing and allocation function according to claim 1, characterized in that, The optimal carbon source dosage will be calculated using gradient descent. Calculate the loss function; ； Where λ is an adjustable parameter. The value of the penalty item. Find the parameter ω that minimizes the loss function, and differentiate the loss function to obtain the following formula: ； ； ； Where α is the learning rate, we find ω and λ when the loss function converges to its minimum, and confirm the weights and bias function; after obtaining the optimal parameters ω, we use ω to calculate the latest total nitrogen, nitrate nitrogen, and influent flow rate, as well as the latest total nitrogen removal rate R. TN Get the latest required carbon source dosage: Among them, R TN This is the latest total nitrogen removal rate.

4. The intelligent carbon source dosing method for wastewater with load sensing and allocation function according to claim 1, characterized in that, It also includes adjusting the amount of carbon source added, which is done through a dynamic learning strategy.

5. A wastewater intelligent carbon source dosing system with load sensing and allocation function, comprising an anoxic tank, a denitrification tank, a carbon source dosing pump unit, an intelligent control cabinet for the carbon source dosing pump, and an intelligent carbon source dosing system; characterized in that, It also includes, The carbon source dosing module adds carbon source to the anoxic tank and denitrification tank through the carbon source dosing pump unit; The carbon source dosing frequency control module controls the carbon source dosing frequency of the carbon source dosing pump unit through the carbon source dosing pump control cabinet; and feeds back the carbon source dosing amount to the carbon source dosing pump control cabinet. The carbon source dosage determination module controls the carbon source dosage by controlling the frequency of the carbon source dosing pump unit through the intelligent carbon source dosing system. The carbon source dosage control module controls the carbon source dosing pump unit through an intelligent carbon source dosing system, thereby controlling the carbon source dosage. The intelligent carbon source dosing algorithm controls the carbon source dosage required for the corresponding anoxic tank load using a machine learning algorithm on the input raw time series dataset. The carbon source dosage required for the total nitrogen removal when using a machine learning algorithm on the input raw time series dataset includes: Among them, R TN ω represents the total amount of nitrogen to be removed, and ω is a model parameter. For the amount of total nitrogen that needs to be removed, R TN The total nitrogen (TN) in the incoming water in (total) and total nitrate nitrogen (NO3) were obtained; ； Among them, TN in (total) represents the total nitrogen content of the incoming water, and NO3(total) represents the total nitrate nitrogen content; for TN in (total) represents the total nitrogen content in the incoming water, measured by the TN content of the incoming water. in and acquisition of incoming water flow; ； Total nitrogen (NO3) is obtained by measuring nitrate nitrogen and influent flow rate. ； Wherein, NO3 represents the amount of nitrate nitrogen; Obtaining the amount of nitrate nitrogen: A model was built based on the total nitrogen in the incoming water and the nitrate nitrogen in the anoxic tank. The model coefficients were obtained using the least squares method, and a real-time calculation method was used to ensure that the model reflects the relationship between total nitrogen and nitrate nitrogen at any given moment. ； Where ℇ represents the difference between the model's predicted value and the actual value; β0, β1, and β2 are the optimal parameters; Fit the model coefficients using the least squares method. To obtain the optimal parameters β0, β1, and β2: By manually setting the target TN for the effluent index target Based on the current inflow index TN in By reverse calculation, the target value of nitrate nitrogen (NO3) in the anoxic pond can be determined. ； Among them, TN target Target values ​​for water discharge indicators are set manually.

Images