AI intelligent liquid caustic soda adding method and system for sewage hardness removal

Abstract

The invention relates to a sewage treatment control technology, and discloses an AI intelligent liquid caustic soda adding method and system for sewage hardness removal, and the method comprises the following steps: adding liquid caustic soda: adding the liquid caustic soda into a hardness removal tank through a liquid caustic soda adding pump unit; the caustic soda liquid adding frequency of the caustic soda liquid adding pump unit is controlled through the caustic soda liquid adding pump control cabinet; the liquid caustic soda adding amount is transmitted to a liquid caustic soda adding pump control cabinet; and the intelligent liquid caustic soda adding system controls the frequency of the liquid caustic soda adding pump unit to control the liquid caustic soda adding amount. According to the invention, the required liquid caustic soda addition amount can be obtained in real time according to the change rule and trend between the input operation index and the equipment regulation and control value, and then the control strategy of liquid caustic soda addition is adjusted according to the liquid caustic soda addition amount. The liquid caustic soda adding amount can be detected in real time, the safety is good, the cost is low, and the operation is simple and convenient.

Description

Technical Field

[0001] The present invention relates to sewage treatment control technology, and in particular to an AI intelligent alkali dosing method and system for sewage hardness removal. Background Art

[0002] In existing sewage treatment systems, hardness removal usually relies on chemical precipitation, a process that mainly involves adding chemical reagents, such as liquid alkali (sodium hydroxide), to wastewater to adjust the pH value, so that calcium and magnesium ions form insoluble precipitates that are removed. However, in traditional operations, the amount of chemical reagents added often relies on manual experience or a simple fixed ratio, resulting in waste or insufficient reagents during the treatment process, which increases operating costs and reduces treatment effects.

[0003] At the same time, due to the large fluctuations in wastewater inflow factors (flow rate, pH, etc.), manual control or traditional automatic control systems cannot adjust the dosage of reagents in real time to cope with changes in inflow, which often leads to unstable treatment effects. Especially in large-scale treatment scenarios, the delay and inaccuracy of manual operations have a negative impact on the overall efficiency and economy of the treatment system.

[0004] Such as the prior art: CN201310527554.2, which discloses a deep sewage treatment system and a treatment method. The structure of the deep sewage treatment system includes an intermediate water tank, a wastewater pump, an oxidation tower, a neutralization and degassing tank, a sedimentation tank, a sludge pump and a sludge concentration tank. The top of the oxidation tower is respectively provided with a concentrated sulfuric acid feed port and a Fenton reagent feed port, the concentrated sulfuric acid feed port is connected to a concentrated sulfuric acid dosing device, and the Fenton reagent feed port is respectively connected to a ferrous sulfate storage tank and a hydrogen peroxide storage tank; the top of the neutralization and degassing tank is provided with a liquid alkali feed port and a coagulant feed port, the liquid alkali feed port is connected to the liquid alkali dosing device, and the coagulant feed port is connected to the coagulant dosing device.

[0005] Although some existing systems have attempted to introduce automated control, most of them are based on preset rules or simple feedback mechanisms, and fail to fully utilize modern AI technology to achieve real-time, dynamic optimization of reagent dosing. These control strategies exhibit problems of delayed response and insufficient control accuracy when faced with complex water quality fluctuations. In addition, existing automated systems generally lack effective use of historical data and cannot achieve long-term learning and optimization.

[0006] Currently, the most similar to the present invention is the automated dosing system based on the feedback control principle. Such systems usually adjust the dosage of liquid alkali in real time by monitoring water quality indicators (such as pH value and hardness). Although the utilization rate of the reagent and the treatment effect have been improved to a certain extent, their control logic is often simple and difficult to cope with complex water quality changes. Most existing technologies cannot achieve comprehensive optimization control based on multiple parameters and multiple variables, resulting in poor dosing accuracy and system stability. In addition, such systems usually require frequent manual correction and debugging, which increases the difficulty of operation and maintenance costs.

[0007] Although some systems have introduced some intelligent control elements, such as using fuzzy control or expert systems for decision-making, their application scope and applicability are still limited. The main problem is that these systems cannot make full use of massive historical data for deep learning, lack adaptability and self-learning capabilities, so that they are still limited to fixed control models in practical applications and cannot be flexibly adjusted with changes in the environment. Summary of the invention

[0008] The present invention aims to address the problems in the prior art that the massive historical data cannot be fully utilized for deep learning, lacks adaptability and self-learning capabilities, and is still limited to fixed control models in practical applications and cannot be flexibly adjusted with environmental changes. The present invention provides an AI intelligent liquid alkali dosing method and system for sewage hardness removal.

[0009] In order to solve the above technical problems, the present invention is solved by the following technical solutions:

[0010] An AI intelligent liquid alkali dosing method for sewage hardness removal includes a hardness removal tank, a liquid alkali dosing pump unit, a liquid alkali dosing pump intelligent control cabinet and an intelligent liquid alkali dosing system; the method includes:

[0011] Liquid alkali is added into the hardness removal tank through the liquid alkali dosing pump unit;

[0012] The control of the liquid alkali dosing frequency is to control the liquid alkali dosing frequency of the liquid alkali dosing pump unit through the liquid alkali dosing pump control cabinet; and the liquid alkali dosing amount is transmitted to the liquid alkali dosing pump control cabinet;

[0013] The amount of liquid alkali to be added is determined by controlling the frequency of the liquid alkali dosing pump unit through the intelligent liquid alkali dosing system.

[0014] Preferably, the implementation of controlling the liquid caustic soda dosage by controlling the frequency of the liquid caustic soda dosing pump unit through the intelligent liquid caustic soda dosing system includes:

[0015] The construction of the AI ​​intelligent caustic soda dosing system database for sewage hardness removal includes the construction of a database for the water inlet data point parameters of the system's hardness removal pool and the liquid caustic soda dosing pump unit data point parameters;

[0016] Preparation of the liquid alkali dosing data set: by preprocessing the input original time series data set, using the pre-order filling principle to fill and correct the abnormal data, the liquid alkali dosing data set is obtained;

[0017] The control of the water inlet of liquid alkali dosing uses deep learning LTSM to predict the time series of incoming water flow data, incoming water pH data, and the target pH value and removal rate value corresponding to complete hardness removal, so as to determine the frequency S' of the next action cycle of the liquid alkali dosing pump.

[0018] As a preferred embodiment, the input original time series data set includes the time series data of pH value of water in the hardness removal pool pH_in, the time series data of water flow in the hardness removal pool Flow_in, the time series data of setting frequency of liquid alkali dosing pump S, and the time series data of pH value of water out of the hardness removal pool pH_out

[0019] Preferably, the system's hardness removal pool inlet data point parameters include inlet flow point data, inlet pH point data, and outlet pH point data; the liquid alkali dosing pump unit data point parameters include liquid alkali dosing pump setting frequency point data, liquid alkali dosing pump feedback frequency point data, and liquid alkali operation status point data.

[0020] Preferably, preprocessing the input original time series data set includes:

[0021] Normalization processing, normalize the water flow data value to the [0,1] interval to facilitate model training. The input data is X t , the normalized data X t 'for:

[0022]

[0023] Among them, X min is the minimum value of the water flow time series data set, X max is the maximum value of the water flow time series data set;

[0024] The water flow time series X is obtained by dividing the data into fixed time windows T to form an input sequence X = [X t-T+1 ,X t-T+2 ,...,X t ];

[0025] Data segmentation, dividing the input sequence X into training set, validation set and test set;

[0026] The formation of the LSTM model, for the segmented data set type LSTM model.

[0027] Preferably, the formation of the LSTM model includes:

[0028] The input layer accepts the preprocessed data sequence X = [X t-T+1 ,X t-T+2 ,...,X t ].

[0029] The LSTM layer is the core of the model and contains multiple memory cells to capture the time dependency in the data. For each time step t, the state update formula of the LSTM cell is as follows:

[0030] i t =σ(W i ·[h t-1 ,x t ]+b i )

[0031] f t =σ(W f ·[h t-1 ,x t ]+b f )

[0032] o t =σ(W o ·[h t-1 ,x t ]+b o )

[0033]

[0034] h t =o t ⊙tanh(C t )

[0035] Among them, i t 、f t and t Respectively represent input gate, forget gate and output gate; C t is the cell state; h t is the hidden state; tanh is the activation function used by LSTM; is the candidate cell state; h t-1 is the hidden state at the previous moment; x t is the input sequence data at the current moment; W i is the weight matrix of the input gate in LSTM. t is the current input, h t-1 is the hidden state at the previous moment, W f is the weight matrix of the forget gate, W o is the weight matrix of the output gate; b i is the bias term of the input gate in LSTM, b f is the bias term of the forget gate, bo is the bias term of the output gate; σ is the sigmoid activation function;

[0036] The fully connected layer converts the output of the LSTM layer into the final prediction result:

[0037]

[0038] Among them, W and b are weight and bias terms, To predict the water flow data at the future time t+1.

[0039] Model optimization, the model training process includes forward propagation and back propagation, the loss function selects the mean square error MSE to evaluate the predictive ability of the model:

[0040]

[0041] Among them, y i is the actual value, is the predicted value.

[0042] Preferably, the target pH value corresponding to complete hardness removal is obtained by:

[0043] H mol =H / N A

[0044] H d =H mol / f

[0045] pH = -log 10 (H d )

[0046] Among them, H mol is the amount of substance of the ion, N A Avogadro's constant is 6.022×10 23 , f is the flow data; calculate the pH value at the corresponding standard flow rate at different flow rates;

[0047] pH is the concentration of H+ in the solution, pH = -log 10 (H + ).

[0048] Preferably, the removal rate r is obtained by:

[0049] The pH value time series data pH_in of the hardness removal pool inlet and the flow rate time series data Flow_in of the hardness removal pool inlet are converted into the pH value time series data pH′ at the corresponding standard flow rate, and the average inlet water pH of the t time series is obtained; the average outlet water pH of the t time series is calculated; the setting frequency S of the liquid alkali dosing pump of the t time series is calculated, and the calculation formula of the removal rate r is as follows:

[0050]

[0051] Among them, S is the set frequency of the liquid alkali dosing pump, and pH′ is the time series data of pH value under the corresponding standard flow rate.

[0052] In order to solve the above technical problems, the present invention also provides an AI intelligent liquid alkali dosing system for sewage hardness removal, including a hardness removal tank, a liquid alkali dosing pump unit, a liquid alkali dosing pump intelligent control cabinet and an intelligent liquid alkali dosing system; it also includes:

[0053] The dosing module of liquid alkali adds liquid alkali into the hardness removal tank through the liquid alkali dosing pump unit;

[0054] The control module of the liquid alkali dosing frequency controls the liquid alkali dosing frequency of the liquid alkali dosing pump unit through the liquid alkali dosing pump control cabinet; and transmits the liquid alkali dosing amount to the liquid alkali dosing pump control cabinet;

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

[0056] The present invention has significant technical effects due to the adoption of the above technical solution:

[0057] The present invention aims at the automatic control algorithm of the liquid alkali addition process in sewage operation, configures the decision model according to the control set value based on the multi-dimensional time series formed by the input multiple operation indicators and the control equipment regulation values, and performs calculation after the algorithm loads the decision model, and then outputs the target regulation value of the control equipment.

[0058] The present invention can obtain the required liquid alkali dosage in real time according to the change law and trend between the input operation index and the equipment control value, and then adjust the liquid alkali dosage control strategy according to the liquid alkali dosage.

[0059] The invention can detect the dosage of liquid alkali in real time and has good safety, low cost and simple operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 is a flow chart of the present invention;

[0061] Figure 2 It is a manual control effect diagram of the prior art;

[0062] Figure 3 It is a control effect diagram of the present invention. DETAILED DESCRIPTION

[0063] The present invention is further described in detail below in conjunction with the accompanying drawings and embodiments.

[0064] Example 1

[0065] An AI intelligent liquid alkali dosing method for sewage hardness removal includes a hardness removal tank, a liquid alkali dosing pump unit, a liquid alkali dosing pump intelligent control cabinet and an intelligent liquid alkali dosing system; the method includes:

[0066] Liquid alkali is added into the hardness removal tank through the liquid alkali dosing pump unit;

[0067] The control of the liquid alkali dosing frequency is to control the liquid alkali dosing frequency of the liquid alkali dosing pump unit through the liquid alkali dosing pump control cabinet; and the liquid alkali dosing amount is transmitted to the liquid alkali dosing pump control cabinet;

[0068] The amount of liquid alkali to be added is determined by controlling the frequency of the liquid alkali dosing pump unit through the intelligent liquid alkali dosing system.

[0069] The realization of controlling the amount of liquid caustic soda by controlling the frequency of the liquid caustic soda dosing pump unit through the intelligent liquid caustic soda dosing system includes:

[0070] The construction of the AI ​​intelligent caustic soda dosing system database for sewage hardness removal includes the construction of a database for the water inlet data point parameters of the system's hardness removal pool and the liquid caustic soda dosing pump unit data point parameters;

[0071] Preparation of the liquid alkali dosing data set: by preprocessing the input original time series data set, using the pre-order filling principle to fill and correct the abnormal data, the liquid alkali dosing data set is obtained;

[0072] The control of the water inlet of liquid alkali dosing uses deep learning LTSM to predict the time series of incoming water flow data, incoming water pH data, and the target pH value and removal rate value corresponding to complete hardness removal, so as to determine the frequency S' of the next action cycle of the liquid alkali dosing pump.

[0073] The input original time series data set includes the pH value time series data of the hardness removal pool inlet water pH_in, the time series data of the hardness removal pool inlet flow rate Flow_in, the time series data of the liquid alkali dosing pump setting frequency S, and the pH value time series data of the hardness removal pool outlet water pH_out

[0074] The system's hardness removal pool inlet data point parameters include inlet flow point data, inlet pH point data, and outlet pH point data; the liquid alkali dosing pump unit data point parameters include liquid alkali dosing pump set frequency point data, liquid alkali dosing pump feedback frequency point data, and liquid alkali operation status point data.

[0075] Preprocessing of the input raw time series dataset includes:

[0076] Normalization processing, normalize the water flow data value to the [0,1] interval to facilitate model training. The input data is X t , the normalized data X t 'for:

[0077]

[0078] Among them, X min is the minimum value of the water flow time series data set, X max is the maximum value of the water flow time series data set;

[0079] The water flow time series X is obtained by dividing the data into fixed time windows T to form an input sequence X = [X t-T+1 ,X t-T+2 ,...,X t ];

[0080] Data segmentation, dividing the input sequence X into training set, validation set and test set;

[0081] The formation of the LSTM model, for the segmented data set type LSTM model.

[0082] Preferably, the formation of the LSTM model includes:

[0083] The input layer accepts the preprocessed data sequence X = [X t-T+1 ,X t-T+2 ,...,X t ].

[0084] The LSTM layer is the core of the model and contains multiple memory cells to capture the time dependency in the data. For each time step t, the state update formula of the LSTM cell is as follows:

[0085] i t =σ(W i ·[h t-1 ,x t ]+b i )

[0086] f t =σ(W f ·[h t-1 ,x t ]+b f )

[0087] o t =σ(W o ·[h t-1 ,x t ]+b o )

[0088]

[0089] h t =o t ⊙tanh(C t )

[0090] Among them, i t 、f t and t Respectively represent input gate, forget gate and output gate; C t is the cell state; h t is the hidden state; tanh is the activation function used by LSTM; is the candidate cell state; h t-1 is the hidden state at the previous moment; x t is the input sequence data at the current moment; W i is the weight matrix of the input gate in LSTM. t is the current input, h t-1 is the hidden state at the previous moment, W f is the weight matrix of the forget gate, W o is the weight matrix of the output gate; b i is the bias term of the input gate in LSTM, b f is the bias term of the forget gate, b o is the bias term of the output gate; σ is the sigmoid activation function;

[0091] The fully connected layer converts the output of the LSTM layer into the final prediction result:

[0092]

[0093] Among them, W and b are weight and bias terms, To predict the water flow data at the future time t+1.

[0094] Model optimization, the model training process includes forward propagation and back propagation, the loss function selects the mean square error MSE to evaluate the predictive ability of the model:

[0095]

[0096] Among them, y i is the actual value, is the predicted value.

[0097] Preferably, the target pH value corresponding to complete hardness removal is obtained by:

[0098] H mol =H / N A

[0099] H d=H mol / f

[0100] pH = -log 10 (H d )

[0101] Among them, H mol is the amount of substance of the ion, N A Avogadro's constant is 6.022×10 23 , f is the flow data; calculate the pH value at the corresponding standard flow rate at different flow rates;

[0102] pH is the concentration of H+ in the solution, pH = -log 10 (H + ).

[0103] Preferably, the removal rate r is obtained by:

[0104] The pH value time series data pH_in of the hardness removal pool inlet and the flow rate time series data Flow_in of the hardness removal pool inlet are converted into the pH value time series data pH′ at the corresponding standard flow rate, and the average inlet water pH of the t time series is obtained; the average outlet water pH of the t time series is calculated; the setting frequency S of the liquid alkali dosing pump of the t time series is calculated, and the calculation formula of the removal rate r is as follows:

[0105]

[0106] Among them, S is the set frequency of the liquid alkali dosing pump, and pH′ is the time series data of pH value under the corresponding standard flow rate.

[0107] Example 2

[0108] An AI intelligent caustic soda dosing system for sewage hardness removal includes a hardness removal tank, a caustic soda dosing pump unit, a caustic soda dosing pump intelligent control cabinet and an intelligent caustic soda dosing system; it also includes:

[0109] The dosing module of liquid alkali adds liquid alkali into the hardness removal tank through the liquid alkali dosing pump unit;

[0110] The control module of the liquid alkali dosing frequency controls the liquid alkali dosing frequency of the liquid alkali dosing pump unit through the liquid alkali dosing pump control cabinet; and transmits the liquid alkali dosing amount to the liquid alkali dosing pump control cabinet;

[0111] The liquid alkali dosage determination module controls the liquid alkali dosage by controlling the frequency of the liquid alkali dosing pump unit through the intelligent liquid alkali dosing system.

[0112] Example 3

[0113] Based on the above embodiments, this embodiment Figure 1 and Figure 2The graphical data shows that when the outlet pH target value is set at 10.90, the mean of the outlet pH of the manual control is 10.74, and the mean of the AI ​​control is 10.92; and the variance of the manual control is 0.04886, and the variance of the AI ​​control is 0.00256; the maximum fluctuation of the manual control is 1.0045, and the maximum fluctuation of the AI ​​control is 0.2020. From the comparison of the control mean, variance, and fluctuation range, the effect of AI control is much better than that of manual control. See the data in Table 1.

[0114] Table 1

[0115]

Claims

1. An AI intelligent caustic soda dosing method for sewage hardness removal, comprising a hardness removal tank, a caustic soda dosing pump unit, a caustic soda dosing pump intelligent control cabinet and an intelligent caustic soda dosing system; The methods include: Liquid alkali is added into the hardness removal tank through the liquid alkali dosing pump unit; The control of the liquid alkali dosing frequency is to control the liquid alkali dosing frequency of the liquid alkali dosing pump unit through the liquid alkali dosing pump control cabinet; and the liquid alkali dosing amount is transmitted to the liquid alkali dosing pump control cabinet; The amount of liquid alkali to be added is determined by controlling the frequency of the liquid alkali dosing pump unit through the intelligent liquid alkali dosing system.

2. The AI ​​intelligent caustic soda dosing method for removing hardness from sewage according to claim 1 is characterized in that: The realization of controlling the amount of liquid caustic soda by controlling the frequency of the liquid caustic soda dosing pump unit through the intelligent liquid caustic soda dosing system includes: The construction of the AI ​​intelligent caustic soda dosing system database for sewage hardness removal includes the construction of a database for the water inlet data point parameters of the system's hardness removal pool and the liquid caustic soda dosing pump unit data point parameters; Preparation of the liquid alkali dosing data set: by preprocessing the input original time series data set, using the pre-order filling principle to fill and correct the abnormal data, the liquid alkali dosing data set is obtained; The control of the water inlet of liquid alkali dosing uses deep learning LTSM to predict the time series of incoming water flow data, incoming water pH data, and the target pH value and removal rate value corresponding to complete hardness removal, so as to determine the frequency S' of the next action cycle of the liquid alkali dosing pump.

3. The AI ​​intelligent caustic soda dosing method for removing hardness from sewage according to claim 2 is characterized in that: The input original time series data set includes the time series data of pH value of water inlet of hardness removal pool pH_in, the time series data of water flow inlet of hardness removal pool Flow_in, the time series data of setting frequency of liquid alkali dosing pump S, and the time series data of pH value of water outlet of hardness removal pool pH_out.

4. The AI ​​intelligent caustic soda dosing method for removing hardness from sewage according to claim 2 is characterized in that: The system's hardness removal pool inlet data point parameters include inlet flow point data, inlet pH point data, and outlet pH point data; the liquid alkali dosing pump unit data point parameters include liquid alkali dosing pump set frequency point data, liquid alkali dosing pump feedback frequency point data, and liquid alkali operation status point data.

5. The AI ​​intelligent caustic soda dosing method for removing hardness from sewage according to claim 2 is characterized in that: Preprocessing of the input raw time series dataset includes: Normalization processing, normalize the water flow data value to the [0,1] interval to facilitate model training. The input data is X t , the normalized data X t 'for: Among them, X min is the minimum value of the water flow time series data set, X max is the maximum value of the water flow time series data set; The water flow time series X is obtained by dividing the data into fixed time windows T to form an input sequence X = [X t-T+1 ,X t-T+2 ,...,X t ]; Data segmentation, dividing the input sequence X into training set, validation set and test set; The formation of the LSTM model, for the segmented data set type LSTM model.

6. The AI ​​intelligent caustic soda dosing method for removing hardness from sewage according to claim 5 is characterized in that: The formation of the LSTM model includes: The input layer accepts the preprocessed data sequence X = [X t-T+1 ,X t-T+2 ,...,X t ]. The LSTM layer is the core of the model and contains multiple memory cells to capture the time dependency in the data. For each time step t, the state update formula of the LSTM cell is as follows: i t =σ(W i ·[h t-1 ,x t ]+b i ) f t =σ(W f ·[h t-1 ,x t ]+b f ) the t =σ(W o ·[h t-1 ,x t ]+b o ) h t =o t ⊙tanh(C t ) Among them, i t 、f t and t Respectively represent input gate, forget gate and output gate; C t is the cell state; h t is the hidden state; tanh is the activation function used by LSTM; is the candidate cell state; h t-1 is the hidden state at the previous moment; x t is the input sequence data at the current moment; W i is the weight matrix of the input gate in LSTM. t is the current input, h t-1 is the hidden state at the previous moment, W f is the weight matrix of the forget gate, W o is the weight matrix of the output gate; b i is the bias term of the input gate in LSTM, b f is the bias term of the forget gate, b o is the bias term of the output gate; σ is the sigmoid activation function; The fully connected layer converts the output of the LSTM layer into the final prediction result: Among them, W and b are weight and bias terms, To predict the water flow data at the future time t+1. Model optimization, the model training process includes forward propagation and back propagation, the loss function selects the mean square error MSE to evaluate the predictive ability of the model: Among them, y i is the actual value, is the predicted value.

7. The AI ​​intelligent caustic soda dosing method for removing hardness from sewage according to claim 2 is characterized in that: The target pH value corresponding to complete hardness removal includes: H mol =H / N A H d =H mol f pH=-log 10 (H d ) Among them, H mol is the amount of substance of the ion, N A Avogadro's constant is 6.022×10 23 , f is the flow data; calculate the pH value at the corresponding standard flow rate at different flow rates; pH is the concentration of H+ in the solution, pH = -log 10 (H + ).

8. The AI ​​intelligent caustic soda dosing method for removing hardness from sewage according to claim 2 is characterized in that: The removal rate r is obtained by: The pH value time series data pH_in of the hardness removal pool inlet and the flow rate time series data Flow_in of the hardness removal pool inlet are converted into the pH value time series data pH′ at the corresponding standard flow rate, and the average inlet water pH of the t time series is obtained; the average outlet water pH of the t time series is calculated; the setting frequency S of the liquid alkali dosing pump of the t time series is calculated, and the calculation formula of the removal rate r is as follows: Among them, S is the set frequency of the liquid alkali dosing pump, and pH′ is the time series data of pH value under the corresponding standard flow rate.

9. An AI intelligent caustic soda dosing system for sewage hardness removal, comprising a hardness removal tank, a caustic soda dosing pump unit, a caustic soda dosing pump intelligent control cabinet and an intelligent caustic soda dosing system; characterized in that: Also includes: The dosing module of liquid alkali adds liquid alkali into the hardness removal tank through the liquid alkali dosing pump unit; The control module of the liquid alkali dosing frequency controls the liquid alkali dosing frequency of the liquid alkali dosing pump unit through the liquid alkali dosing pump control cabinet; and transmits the liquid alkali dosing amount to the liquid alkali dosing pump control cabinet; The liquid alkali dosage determination module controls the liquid alkali dosage by controlling the frequency of the liquid alkali dosing pump unit through the intelligent liquid alkali dosing system.

Images