A control method and device of an ultrapure water preparation system, a medium and an electronic device

Abstract

The application relates to the field of ultrapure water preparation and provides a control method and device for an ultrapure water preparation system, a medium and electronic equipment, aiming to improve prediction accuracy and the accuracy of recovery rate adjustment. The method comprises inputting time series data into a first target prediction model to generate scaling rate information of a future period, wherein the time series data comprises RO system water inlet parameter characteristics, system working configuration parameter sequences and pre-water treatment system performance parameters. According to the scaling rate information, a recovery rate adjustment range is generated, and the real-time recovery rate of the RO system is dynamically controlled within the adjustment period. Through prediction model and pre-water treatment system performance prediction, the overall prediction accuracy is improved, and the influence of the pre-water treatment system on the RO system is considered, thereby providing more accurate prediction results. The problems of hysteresis and low adjustment accuracy in the prior art are solved, the water treatment efficiency is improved, and the possibility of RO system scaling is reduced through dynamic control of the recovery rate.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of ultrapure water preparation, and particularly to a control method and device for an ultrapure water preparation system, a medium, and an electronic device. BACKGROUND

[0002] Ultrapure water is high-purity water that has almost all impurities and ions removed, and its conductivity is close to the theoretical limit value (0.055 μS / cm, corresponding to a resistivity of 18.2 MΩ·cm). It is widely used in fields such as semiconductor manufacturing, biomedicine, and analytical testing, which have very high requirements for water quality.

[0003] The basic process of ultrapure water preparation mainly includes four stages: pretreatment, primary desalination, deep desalination and purification, and terminal treatment and storage. The goal of pretreatment is to remove large-particle impurities, residual chlorine, organic matter, and microorganisms in raw water to protect subsequent treatment equipment. Common methods include a multi-media filter (for removing suspended solids and colloids), activated carbon adsorption (for removing residual chlorine and organic matter), a softener (for removing calcium and magnesium ions to prevent scaling by ion exchange resin), and a precision filter (for intercepting small particles). There is also ultrafiltration (UF) for intercepting macromolecular substances such as proteins, viruses, and colloids. Primary desalination aims to remove most of the dissolved inorganic salts. Main technologies include reverse osmosis (RO), which uses a semi-permeable membrane to remove 95% to 99% of ions, organic matter, and microorganisms under high pressure; and nanofiltration (NF), which is suitable for removing divalent ions and macromolecular organic matter. Subsequent deep desalination and purification are used to further remove residual ions and trace impurities to make the water quality meet the standard of ultrapure water. Main technologies include electrodeionization (EDI), mixed bed ion exchange, and ultraviolet sterilization (UV) treatment processes.

[0004] Reverse osmosis (RO) technology is a high-efficiency desalination method that uses a semi-permeable membrane to separate water and solutes under pressure. The water recovery rate (also known as the recovery rate or concentration ratio) of an RO system, which is the ratio of product water flow to feed water flow, is one of the important indicators for measuring the efficiency of an RO system. There is also a negative correlation between the water recovery rate and the risk of scaling. When the water recovery rate of an RO system increases, it means that more water is converted into product water, and less water flows through the membrane surface as concentrated water. This will cause the ion concentration on the concentrated water side to increase significantly, especially for salts with low solubility (such as calcium sulfate and calcium carbonate), which are more likely to reach the saturation point and start to precipitate and form scaling. Therefore, a higher water recovery rate usually comes with a higher risk of scaling. Conversely, if the water recovery rate is reduced, the amount of concentrated water discharged will increase, thereby reducing the ion concentration on the concentrated water side, making it more difficult for various salts to reach their saturation concentrations, and reducing the likelihood of scaling. However, doing so will reduce the overall net water efficiency of the system.

[0005] In the prior art, the recovery rate of the RO system is usually adjusted by PID (Proportional-Integral-Derivative Controller) control, similar to a feedback mechanism, such as dynamically adjusting the recovery rate according to the ion concentration data on the concentrated water side, but this adjustment usually has a certain lag after a certain phenomenon occurs. Moreover, the adjustment range is usually determined by humans, so that the adjustment accuracy of the recovery rate is low. SUMMARY

[0006] To solve one of the above technical problems, the present application adopts the technical solution of:

[0007] According to one aspect of the present application, a control method for an ultrapure water preparation system is provided, the method comprising the following steps:

[0008] inputting time series data corresponding to the current adjustment time into a first target prediction model to generate scaling rate information of a future period corresponding to the current adjustment time; the time series data comprising water inlet parameter features of the RO system in a historical period corresponding to the current adjustment time, system working configuration parameter sequences of the RO system corresponding to the future period, and water treatment performance parameters of the preposed water treatment system corresponding to a preposed future period; the preposed water treatment system being a water treatment system located before the RO system in the process flow sequence; the preposed future period being a period in which the preposed water treatment system affects the RO system in the future period;

[0009] generating a recovery rate adjustment range of the RO system corresponding to the current adjustment time according to the scaling rate information of the future period corresponding to the current adjustment time;

[0010] controlling the real-time recovery rate of the RO system to dynamically change within the recovery rate adjustment range in an adjustment period corresponding to the current adjustment time;

[0011] The water treatment performance parameters of the preposed water treatment system corresponding to the future period are obtained according to the following steps:

[0012] obtaining a water inlet feature sequence of the preposed water treatment system in a historical period corresponding to the current adjustment time;

[0013] obtaining system working configuration parameter sequences corresponding to the historical period corresponding to the current adjustment time and / or the preposed future period of the preposed water treatment system, respectively;

[0014] inputting the water inlet feature sequence and the system working configuration parameter sequences of the preposed water treatment system corresponding to the current adjustment time into a second target prediction model to generate the water treatment performance parameters of the preposed water treatment system corresponding to the preposed future period.

[0015] According to a second aspect of the present application, a control device of an ultrapure water preparation system is provided, comprising:

[0016] a prediction module configured to input time series data corresponding to the current adjustment time into a first target prediction model to generate scaling rate information of a future period corresponding to the current adjustment time, wherein the time series data comprises water inlet parameter characteristics of the RO system in a historical period corresponding to the current adjustment time, a system working configuration parameter sequence of the RO system in the future period, and water treatment performance parameters of a pre-water treatment system in a pre-future period; the pre-water treatment system is a water treatment system located before the RO system in a process flow sequence; the pre-future period is a period in which the pre-water treatment system has an impact on the RO system in the future period;

[0017] a recovery rate generation module configured to generate a recovery rate adjustment range of the RO system corresponding to the current adjustment time according to the scaling rate information of the future period corresponding to the current adjustment time;

[0018] an adjustment control module configured to control the real-time recovery rate of the RO system to dynamically change within the recovery rate adjustment range in an adjustment period corresponding to the current adjustment time;

[0019] The prediction module is further configured to obtain the water treatment performance parameters of the pre-water treatment system in the future period according to the following steps:

[0020] obtain a water inlet characteristic sequence of the pre-water treatment system in a historical period corresponding to the current adjustment time;

[0021] obtain a system working configuration parameter sequence corresponding to the historical period corresponding to the current adjustment time and / or the pre-future period;

[0022] input the water inlet characteristic sequence and the system working configuration parameter sequence of the pre-water treatment system corresponding to the current adjustment time into a second target prediction model to generate the water treatment performance parameters of the pre-water treatment system in the pre-future period.

[0023] According to a third aspect of the present application, a non-transitory computer readable storage medium is provided, which stores a computer program, and the computer program is executed by a processor to implement the control method of the ultrapure water preparation system.

[0024] According to a fourth aspect of the present application, an electronic device is provided, which comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the control method of the ultrapure water preparation system when executing the computer program.

[0025] The present application has at least one of the following beneficial effects:

[0026] In the prior art, the PID control mechanism has a lag, and the adjustment range is usually determined by humans, resulting in low adjustment accuracy of the recovery rate. The present application uses the first target prediction model and the second target prediction model to predict the performance of the pre-treatment system to assist in predicting changes in water quality in the later stage, thereby improving the overall prediction accuracy. In the prediction process, various parameters of the pre-treatment system (such as the incoming water characteristic sequence and the system working configuration parameter sequence) are considered comprehensively and input into the second target prediction model to generate water treatment performance parameters in the future period. This method can more comprehensively reflect the influence of the pre-treatment system on the RO system, thereby providing more accurate prediction results.

[0027] At the same time, by predicting the scaling rate information in advance and generating a recovery rate adjustment range, especially by adding a limiting mechanism for the upper limit of the recovery rate, the system can make more accurate and advance recovery rate adjustments. The problem of lag and low adjustment accuracy in the prior art is solved. In addition, by dynamically controlling the real-time recovery rate of the RO system to change within the recovery rate adjustment range, the water treatment efficiency can be improved while the possibility of scaling of the RO system is reduced as much as possible. BRIEF DESCRIPTION OF DRAWINGS

[0028] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0029] Figure 1 A flow chart of a control method of an ultrapure water preparation system provided by an embodiment of the present application is shown in the figure.

[0030] Figure 2 A structural schematic diagram of a control device of an ultrapure water preparation system provided by an embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0031] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings of the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0032] As a possible embodiment of the present application, as shown in the figure, a control method of an ultrapure water preparation system is provided, which comprises the following steps: Figure 1

[0033] ​S100: input the time series data corresponding to the current adjustment time into the first target prediction model to generate the scaling rate information of the future period corresponding to the current adjustment time. The time series data includes the influent parameter characteristics of the RO system in the historical period corresponding to the current adjustment time, the system working configuration parameter sequence of the RO system in the future period, and the water treatment performance parameters of the pre-water treatment system in the pre-future period. The pre-water treatment system is a water treatment system located before the RO system in the process flow sequence. The pre-future period is a period in which the pre-water treatment system affects the RO system in the future period.

[0034] Specifically, in the embodiment, the first target prediction model and the first target prediction model can both be LSTM (Long Short-Term Memory networks) models. The adjustment period can be two hours, and the adjustment time of each adjustment period is the starting time of the adjustment period. The length of the future period is greater than the length of the adjustment period.

[0035] In the embodiment, taking the adjustment period of two hours, the future period of 6 hours, and the historical period of 12 hours as an example, the operation process of the scheme of the embodiment is described as follows:

[0036] At the starting time of each adjustment period (such as 0-2 hours), the system collects the time series data of the past 12 hours (historical period) corresponding to the current adjustment time, inputs the first target prediction model, and predicts the scaling rate information of the future 6 hours (future period). That is, the system recalculates the RO system recovery rate adjustment range every two hours based on the current data, and dynamically adjusts the recovery rate in the period. The adjustment period determines the system update frequency, which can be manually set according to needs. In the example, the short interval of two hours can achieve high-frequency dynamic adjustment and reduce the risk of scaling.

[0037] In addition, because the quality of the influent water treated in the historical period will affect the filtering performance of the RO system, and in turn affect the effect of the purified water in the future period, the future state can be predicted based on the time series data.

[0038] The pre-water treatment system can be one or more. If the pre-water treatment system is multiple, the water treatment performance parameters of the pre-water treatment system in the future period are configured with corresponding parameter adjustment weights according to the process distance between the pre-water treatment system and the RO system. The parameter adjustment weight is negatively correlated with the process distance.

[0039] Since there are also multiple process steps between the RO systems, the devices in these steps treat the current influent through respective treatment mechanisms to remove corresponding impurities, and the performance of the water treatment systems located before the RO systems also affects the treatment effect and the scaling risk of the RO systems, and the closer the water treatment system to the RO system, the greater the impact on the RO system. Therefore, when setting the corresponding water treatment performance parameters, the smaller the process distance between the pre-water treatment system and the RO system, the higher the parameter adjustment weight of the corresponding configuration, so that the model better considers the influence of the water treatment system on the RO system in calculation.

[0040] The pre-water treatment system in this embodiment includes any one or several of a multi-medium filter, an activated carbon adsorption device, a softener, and a precision filter.

[0041] Specifically, the influent characteristics, system working configuration parameters, and water treatment performance parameters of each of the above pre-water treatment systems can be set as follows:

[0042] The influent characteristics of the multi-medium filter include suspended solids and colloidal particle content. The system working configuration parameters include filter material type, filtration speed, filter core usage rate, and backwash frequency and intensity. The water treatment performance parameters include suspended solids removal rate and turbidity reduction rate.

[0043] The influent characteristics of the activated carbon adsorption device include residual chlorine and organic matter content. The system working configuration parameters include activated carbon type, filling amount, contact time, filter core usage rate, and water flow rate. The water treatment performance parameters include residual chlorine removal rate and total organic carbon removal rate.

[0044] The influent characteristics of the softener include calcium and magnesium ion concentration. The system working configuration parameters include ion exchange resin type, ion exchange resin regeneration period, and salt consumption. The water treatment performance parameters include hardness removal rate and effluent hardness value.

[0045] The influent characteristics of the precision filter include micro-particle content. The system working configuration parameters include filter core precision, flow rate, filter core usage rate, and replacement period. The water treatment performance parameters include particle retention rate and effluent turbidity.

[0046] The water treatment performance parameters of the pre-water treatment system in the future period are obtained according to the following steps:

[0047] S101: Obtain the influent characteristic sequence of the pre-water treatment system in the corresponding historical period at the current adjustment time.

[0048] S102: Obtain the system working configuration parameter sequence corresponding to the historical period and / or the pre-future period at the current adjustment time of the pre-water treatment system, respectively.

[0049] In this step, because the process distance between each pre-water treatment system and the RO system is different, the pre-future time period corresponding to each pre-water treatment system is different. For example, the process sequence from front to back is a multi-medium filter, an activated carbon adsorption device, a softener, and an RO system. Then, the water treated by the multi-medium filter reaches the RO system, that is, the RO system is affected, and there is a certain hysteresis. And because the process distance is different, the hysteresis time length corresponding to different pre-water treatment systems is also different. For example, the hysteresis time length of the multi-medium filter is 10 minutes, the hysteresis time length of the softener is 5 minutes, the future time period corresponding to the current adjustment time is 10:10:00 to 10:20:00, then the pre-future time period corresponding to the multi-medium filter is 10:00:00 to 10:10:00, and the pre-future time period corresponding to the softener is 10:05:00 to 10:15:00.

[0050] Because some pre-water treatment systems will keep the working parameters unchanged for a long time within the prediction period, the system working configuration parameter sequence corresponding to the historical time period or the pre-future time period corresponding to the current adjustment time can be used, and because some pre-water treatment systems will change the working parameters within the prediction period, in order to ensure the prediction accuracy, the system working configuration parameter sequence corresponding to the historical time period and the pre-future time period corresponding to the current adjustment time is needed. Usually, the change of the working parameters is planned in advance according to the use, so it can be obtained according to the use plan.

[0051] S103: Input the water inlet characteristic sequence and the system working configuration parameter sequence of the pre-water treatment system corresponding to the current adjustment time into the second target prediction model, and generate the water treatment performance parameter of the pre-water treatment system corresponding to the pre-future time period.

[0052] In this embodiment, the training sample can be generated according to the historical data to train the LSTM model, so as to obtain the corresponding first target model and second target model.

[0053] In this embodiment, the first target model and the second target model are both existing LSTM models. The training steps of the model can be determined according to the existing technology, and the training process of the LSTM model is described as follows.

[0054] Data preparation:

[0055] a. Collect and organize historical data to generate training data sets, and ensure that the data quality meets the model training requirements.

[0056] b. Preprocess the data, including normalization, denoising, missing value filling, etc.

[0057] c. Split the dataset into training, validation, and test sets.

[0058] Input form:

[0059] a. For time series data, the input form is sequential data, i.e., each sample consists of a series of observations at different time points.

[0060] b. For non-time series data, the input form is a fixed-length feature vector.

[0061] Training labels:

[0062] a. For the first target prediction model, the label is the future period's RO system fouling rate information corresponding to the current adjustment time.

[0063] b. For the second target model, the label is the water treatment performance parameter corresponding to the pre-future period of the pre-treatment system.

[0064] In this embodiment, the input form and label of the training sample corresponding to the prediction requirements of the two models are illustrated with the following examples:

[0065] The input form of the first target prediction model (RO system fouling rate prediction):

[0066] Contains three parts of time series data:

[0067] Historical period influent parameter features of the RO system (X1, X2,..., Xn): including conductivity, pH value, temperature, suspended solids concentration, and other parameters in the past 12 hours of historical time series data.

[0068] Future period working configuration parameter sequence of the RO system (Y1, Y2,..., Yz): including membrane element type, operating pressure, influent flow, and other future 6-hour preset parameters.

[0069] Future period performance parameters of the pre-treatment system (W11, W12,..., W2q): predicted output from the second target model, such as suspended solids removal rate, residual chlorine removal rate, etc.

[0070] The input data dimension includes historical data statistical features (mean / extreme value / trend change) and time series sequence of future configuration parameters.

[0071] Training labels:

[0072] The output is the future period's fouling rate information:

[0073] The fouling probability sequence is the fouling probability value of each sub-period (e.g., each hour) of the future period; or the fouling probability change curve is a predicted curve reflecting the change of the fouling probability over time.

[0074] The label data is generated by actual fouling events in the historical operation data.

[0075] The acquisition of the fouling rate information is illustrated as follows:

[0076] It can be determined in the historical data when the RO system is fouled and when it is not fouled, and the change in the fouling degree is usually not particularly rapid, so it can be set according to the change of the fouling degree within one hour. In this embodiment, the fouling degree can be taken as the fouling rate value. In addition, the fouling degree is usually in a gradual change process. Therefore, in this embodiment, different fouling rate values can be assigned to each hour within the period from the use of the new permeable membrane to the fouling of the permeable membrane by expert experience, and in addition, the training data needs to be acquired within each use cycle of the permeable membrane. For example, when the new permeable membrane is used for 30 hours, the fouling degree reaches a state that cannot be used and needs to be replaced. The fouling probability values of each hour within these 30 hours can be configured as follows: the fouling probability values corresponding to [1, 10] hours are all 0, the fouling probability values corresponding to each hour within [11, 20] hours are all 0.04×(Td-10); Td is the hour number corresponding to the current time, and the fouling probability values corresponding to each hour within [21, 30] hours are all 0.06×(Td-20). Of course, this example is only an example for illustration, and specifically, the fouling rate information of each period can be determined according to the corresponding fouling degree. Therefore, the fouling probability value sequence corresponding to every 6 hours can be used as the corresponding label.

[0077] For example, if the adjustment time belongs to the 11th hour of the use of the new permeable membrane, the fouling rate values of the corresponding future 6 hours are the fouling rate values of [12, 17] hours, i.e., 0.08, 0.12, 0.16, 0.2, 0.24, and 0.28.

[0078] The input form of the second target prediction model (the pre-treatment system performance prediction):

[0079] It contains two parts of data:

[0080] The influent feature sequence (D1, D2,..., Dp) of the pre-treatment water treatment system: such as the historical influent parameter time series data of suspended solids content, residual chlorine concentration, etc.

[0081] The configuration parameter sequence (E1, E2,..., Eh) of the pre-treatment system: including the filter material type, the historical period and / or the configuration parameters of the pre-treatment future period, etc.

[0082] Particularly consider process distance weight: assign adjustment weight to multiple pre-systems according to process distance with RO system.

[0083] Training labels:

[0084] Output is water treatment performance parameter of pre-system in future period: specific indicators are determined according to device type (such as turbidity reduction rate of multi-medium filter, hardness removal rate of softener, etc.), and its form can include performance parameter prediction value in time series form (such as hourly removal rate within 6 hours). The water treatment performance parameters in this embodiment can be obtained from the performance monitoring device configured by the existing equipment, and then the sequence value corresponding to each hour is calculated into the mean value, or the mode, median and other statistical characteristic values are taken as the water treatment performance parameter corresponding to the hour.

[0085] Model initialization:

[0086] a. Initialize the weights and biases of the LSTM model.

[0087] b. Set appropriate hyperparameters, such as learning rate, batch size, number of iterations, etc.

[0088] Training process:

[0089] a. Use the training set data to perform supervised training on the model.

[0090] b. In each training cycle (epoch), input the input sequence and corresponding label into the model.

[0091] c. Calculate the loss function, such as mean square error (MSE) or cross-entropy loss.

[0092] d. Update the model weights through the backpropagation algorithm.

[0093] e. Evaluate the model performance using the validation set, and adjust the hyperparameters to prevent overfitting.

[0094] Model evaluation and optimization:

[0095] a. Evaluate the model performance on the validation set, monitor the loss function value and evaluation indicators.

[0096] b. Adjust the model structure or hyperparameters according to the validation results.

[0097] c. Repeat the training and evaluation steps until the model performance reaches the predetermined standard or no longer improves.

[0098] Since the model in training and subsequent prediction use, its input data form is exactly the same, in the embodiment, taking the current adjustment time as 24 days 00:00:00, the adjustment period as two hours, the future period as 6 hours, the history period as 12 hours, and the pre-water treatment system as two, the input data form of the first target model and the second target model in the embodiment is described as follows:

[0099] The input data form of the first target model is as follows:

[0100] X1, X2, …, Xn, Y1, Y2, …, Yz, W 1 1, W 1 2, …, W 1 m, W 2 1, W 2 2, …, W 2 q, wherein X1, X2, …, Xn respectively represent n water inlet parameter characteristics of the RO system, and time sequence data sequences respectively collected in the past 12 hours (23 days 12:00:00 to 24:00:00). The water inlet parameter characteristics can include conductivity (μS / cm), pH value, temperature (℃), suspended matter concentration (mg / L), colloidal particle size distribution (μm), total residual chlorine concentration (ppm), and total organic carbon TOC value (ppb), hardness index after softener treatment, expressed in CaCO3 equivalent (mg / L), and micro-particle content, represented by turbidity (NTU) or particle counter (pieces / mL). The time sequence data corresponding to the water inlet parameter characteristics can be composed of the average value of the water inlet parameter characteristics in each fixed time period, and the maximum value, minimum value, trend change value and other characteristic values in the entire history period (23 days 12:00:00 to 24:00:00).

[0101] Y1, Y2, …, Yz respectively represent z system working configuration parameter sequences of the RO system in the future period (24 days 00:00:00 to 06:00:00), and the system working configuration parameters can include membrane element type, operating pressure, water inlet flow, concentrated water discharge, temperature control parameter, pH adjuster dosage, chemical cleaning cycle (such as weekly acid washing / monthly alkali washing), physical cleaning frequency (such as backwashing interval), and membrane module series / parallel mode. These system working configuration parameters can also generate time sequence data of each system working configuration parameter corresponding to the future period according to the processing method of the time sequence data of the water inlet parameter characteristics.

[0102] W 1 1, W 1 2, …, W 1 m, W 21、W 2 2、……、W 2 q represents q water treatment performance parameters corresponding to the future period of another pre-water treatment system, which can be the average value of the entire future period, or can generate time series data of each water treatment performance parameter corresponding to the future period according to the processing mode of the time series data of the influent parameter characteristics.

[0103] The input data form of the second target model is as follows:

[0104] D1, D2, …, Dp, E1, E2, …, Eh, wherein D1, D2, …, Dp are p influent characteristic sequences of the pre-water treatment system corresponding to the current adjustment time, and the acquisition method can refer to the influent characteristic sequence of the RO system. E1, E2, …, Eh are h system working configuration parameters of the pre-water treatment system corresponding to the pre-future period. The specific pre-future period can be determined according to the lag between the specific RO system and the pre-water treatment system.

[0105] S200: generating the recovery rate adjustment range of the RO system corresponding to the current adjustment time according to the scaling rate information of the future period corresponding to the current adjustment time.

[0106] S200 can be implemented in the following two ways according to different scaling rate information.

[0107] First, the scaling rate information includes scaling probabilities corresponding to multiple sub-periods in the future period.

[0108] S200 includes:

[0109] S201: generating an upper limit value Hmax of the recovery rate adjustment range according to the scaling probabilities corresponding to the multiple sub-periods. Hmax satisfies the following conditions:

[0110] .

[0111] Wherein, Q1 is a preset recovery rate of the RO system. p i and a i are the scaling probability and weight corresponding to the i-th sub-period in the future period, respectively. The weight corresponding to the sub-period is negatively related to the order. K is a gain adjustment coefficient. Q2 is a preset maximum recovery rate of the RO system, and Hmin < Q1 < Q2. Hmin is the lower limit value of the recovery rate adjustment range.

[0112] For example, the scaling rate information is the scaling probability corresponding to each hour in the future 6 hours. Therefore, the scaling rate adjustment floating value can be generated according to the scaling probability of each hour, that is, Thus, the size of Hmax can be dynamically adjusted according to the predicted plurality of scaling probabilities, specifically, if is greater, Hmax is smaller, and if is smaller, Hmax is greater, so that Hmax is more in line with the future scaling rate risk limit.

[0113] In addition, the lower limit value Hmin of the recovery rate adjustment range in the present embodiment can be determined by the minimum water treatment efficiency currently required by the RO system.

[0114] Secondly, the scaling rate information includes a scaling probability change curve in the future period.

[0115] S200 includes:

[0116] S202: According to the scaling probability change curve, the recovery rate adjustment range of the RO system corresponding to the current adjustment time is selected.

[0117] In the present embodiment, the recovery rate adjustment range of the RO system corresponding to the current adjustment time can be selected by a person according to the trend represented by the scaling probability change curve.

[0118] S300: In the current adjustment period corresponding to the current adjustment time, the real-time recovery rate of the RO system is controlled to dynamically change within the recovery rate adjustment range.

[0119] With the above more accurate recovery rate adjustment range limit, the existing recovery rate adjustment method (such as PID or manual adjustment according to experience) can be adjusted within a more suitable and accurate range, and the situation of adjusting beyond the reasonable range can be avoided as much as possible.

[0120] As another possible embodiment of the present application, as Figure 2 shown, a control device of an ultrapure water preparation system is also provided, comprising:

[0121] A prediction module is configured to input time series data corresponding to the current adjustment time into a first target prediction model to generate scaling rate information of a future period corresponding to the current adjustment time. The time series data includes water inlet parameter features of the RO system in a historical period corresponding to the current adjustment time, a system working configuration parameter sequence of the RO system in the future period, and water treatment performance parameters of a preposed water treatment system in a preposed future period. The preposed water treatment system is a water treatment system located before the RO system in the process flow sequence. The preposed future period is a period in which the preposed water treatment system affects the RO system in the future period.

[0122] A recovery rate generation module is configured to generate a recovery rate adjustment range of the RO system corresponding to the current adjustment time according to the scaling rate information of the future period corresponding to the current adjustment time.

[0123] The adjustment control module is configured to control the real-time recovery rate of the RO system to dynamically change within the recovery rate adjustment range during the corresponding adjustment period at the current adjustment moment.

[0124] The prediction module is further configured to obtain the water treatment performance parameter of the pre-arranged water treatment system corresponding to the future period according to the following steps.

[0125] The pre-arranged water treatment system is obtained at the current adjustment moment corresponding to the historical period of the water inlet characteristic sequence.

[0126] The pre-arranged water treatment system is obtained at the current adjustment moment corresponding to the historical period and / or the pre-arranged future period of the system working configuration parameter sequence.

[0127] The pre-arranged water treatment system is obtained at the current adjustment moment corresponding to the water inlet characteristic sequence and the system working configuration parameter sequence, and the water treatment performance parameter of the pre-arranged water treatment system corresponding to the pre-arranged future period is generated.

[0128] In addition, although the various steps of the method in the present disclosure are described in a specific order in the accompanying drawings, this does not require or imply that the steps must be performed in this specific order, or that all of the steps shown must be performed to achieve the desired results. Additionally or alternatively, certain steps can be omitted, multiple steps can be combined into one step, and / or one step can be divided into multiple steps, etc.

[0129] From the above description of the embodiments, those skilled in the art can easily understand that the example embodiments described herein can be implemented by software, or by software in combination with necessary hardware. Therefore, the technical solutions according to the embodiments of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a U disk, a mobile hard disk, etc.) or a network, and includes a number of instructions to make a computing device (which can be a personal computer, a server, a mobile terminal, or a network device, etc.) execute the method according to the embodiments of the present disclosure.

[0130] In the exemplary embodiments of the present disclosure, an electronic device capable of implementing the above method is also provided.

[0131] Those skilled in the art can understand that each aspect of the present disclosure can be implemented as a system, a method or a program product. Therefore, each aspect of the present disclosure can be embodied in the form of a complete hardware embodiment, a complete software embodiment (including firmware, microcode, etc.), or an embodiment combining hardware and software aspects, which can be collectively referred to as "circuitry", "module" or "system" herein.

[0132] The electronic device according to this embodiment of the present application. The electronic device is merely an example and should not bring any limitation to the function and use range of the embodiments of the present application.

[0133] The electronic device is in the form of a general purpose computing device. The components of the electronic device can include, but are not limited to, the at least one processor described above, the at least one memory described above, a bus that connects different system components (including the memory and the processor).

[0134] The memory stores program codes which can be executed by the processor, so that the processor executes the steps described in the above "Exemplary Method" section according to various exemplary embodiments of the present application.

[0135] The memory can include a readable medium in the form of a volatile memory, such as a random access memory (RAM) and / or a cache memory, and can further include a read only memory (ROM).

[0136] The memory can further include programs / utilities with a set of (at least one) program modules, such as an operating system, one or more application programs, other program modules, and program data, each of which or some combination of which can include implementation of a network environment.

[0137] The bus can be one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, a graphics acceleration bus, a processor or local bus using any of a variety of bus structures, and the like.

[0138] The electronic device can also communicate with one or more external devices (such as a keyboard or a pointing device, a Bluetooth device, etc.) and can also communicate with one or more devices that enable a user to interact with the electronic device (and / or one or more input / output (I / O) devices 620). Furthermore, the electronic device can communicate with one or more networks (such as a local area network (LAN), a wide area network (WAN), and / or the public network, such as the Internet) through a network adapter. The network adapter communicates with the other modules of the electronic device via the bus. It should be appreciated that although not shown, other hardware and / or software modules could be used in conjunction with the electronic device. Such hardware and / or software modules can include, but are not limited to, microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data archival storage systems, etc.

[0139] Those skilled in the art can clearly understand, through the description of the above embodiments, that the example embodiments described herein can be implemented by software, or by software in combination with necessary hardware. Therefore, the technical solutions according to the embodiments of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a U disk, a mobile hard disk, etc.) or a network, and includes a plurality of instructions to enable a computing device (which can be a personal computer, a server, a terminal device, or a network device, etc.) to perform the method according to the embodiments of the present disclosure.

[0140] In the example embodiments of the present disclosure, a computer readable storage medium is also provided, which stores a program product capable of implementing the above-mentioned method of the present disclosure. In some possible embodiments, various aspects of the present disclosure can also be implemented in the form of a program product, which includes program code for causing a terminal device to perform the steps according to various example embodiments of the present disclosure described in the above-mentioned “example method” section of the present disclosure when the program product is run on the terminal device.

[0141] The program product can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium may, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or component, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include an electrical connection having one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0142] The computer readable signal medium can include a data signal propagated in a baseband or as a part of a carrier wave, in which readable program code is borne. Such a propagated data signal can take on multiple forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above. The readable signal medium can also be any readable medium other than the readable storage medium, which can send, propagate or transmit the program for use by or in connection with an instruction execution system, device or component.

[0143] The program code contained on the readable medium can be transmitted by any suitable medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination of the above.

[0144] Program code for performing the operations of this invention can be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java and C++, and conventional procedural programming languages ​​such as C or similar languages. The program code can execute entirely on the user's computing device, partially on the user's device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0145] Furthermore, the above figures are merely illustrative of the processes included in the method according to exemplary embodiments of the present invention, and are not intended to be limiting. It is readily understood that the processes shown in the above figures do not indicate or limit the temporal order of these processes. Additionally, it is readily understood that these processes may be executed synchronously or asynchronously, for example, in multiple modules.

[0146] It should be noted that although several modules or units for the device used to perform actions have been mentioned in the detailed description above, this division is not mandatory. In fact, according to embodiments of this disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0147] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A control method of an ultrapure water preparation system, characterized by, The method comprises the following steps: inputting time sequence data corresponding to the current adjustment time into a first target prediction model to generate scale rate information of a future period corresponding to the current adjustment time; the time sequence data comprises water inlet parameter characteristics of the RO system in a historical period corresponding to the current adjustment time, a system working configuration parameter sequence of the RO system corresponding to the future period, and water treatment performance parameters of the pre-treatment water treatment system corresponding to a pre-treatment future period; the pre-treatment water treatment system is a water treatment system located before the RO system in the process sequence; the pre-treatment future period is a period in which the pre-treatment water treatment system affects the RO system in the future period; generating a recovery rate adjustment range of the RO system corresponding to the current adjustment time according to the scale rate information of the future period corresponding to the current adjustment time; controlling the real-time recovery rate of the RO system to dynamically change within the recovery rate adjustment range in an adjustment period corresponding to the current adjustment time; the water treatment performance parameters of the pre-treatment water treatment system corresponding to the future period are obtained by the following steps: obtaining a water inlet characteristic sequence of the pre-treatment water treatment system in a historical period corresponding to the current adjustment time; obtaining a system working configuration parameter sequence corresponding to the historical period corresponding to the current adjustment time and / or the pre-treatment future period; inputting the water inlet characteristic sequence and the system working configuration parameter sequence of the pre-treatment water treatment system corresponding to the current adjustment time into a second target prediction model to generate the water treatment performance parameters of the pre-treatment water treatment system corresponding to the pre-treatment future period; both the first target prediction model and the second target prediction model are LSTM models.

2. The method of claim 1, wherein, the scale rate information comprises scale probabilities corresponding to multiple sub-periods in the future period; generating a recovery rate adjustment range of the RO system corresponding to the current adjustment time according to the scale rate information of the future period corresponding to the current adjustment time comprises: generating an upper limit value Hmax of the recovery rate adjustment range according to the scale probabilities corresponding to the multiple sub-periods; Hmax satisfies the following condition: ； wherein Q1 is a preset recovery rate conventional value of the RO system; p i and a i are the fouling probability and weight of the i-th sub-period in the future period, respectively, the weight of the sub-period is negatively related to the order; K is a gain adjustment coefficient; Q2 is a preset recovery rate maximum value of the RO system, Hmin < Q1 < Q2; Hmin is a lower limit value of the recovery rate adjustment range.

3. The method of claim 1, wherein, the scale rate information comprises a scale probability change curve in the future period; generating a recovery rate adjustment range of the RO system corresponding to the current adjustment time according to the scale rate information of the future period corresponding to the current adjustment time comprises: selecting the recovery rate adjustment range of the RO system corresponding to the current adjustment time according to the scale probability change curve.

4. The method of claim 1, wherein, the adjustment period is two hours, the adjustment time of each adjustment period is the starting time of the adjustment period, and the length of the future period is greater than the length of the adjustment period.

5. The method of claim 1, wherein, the pre-treatment water treatment system comprises any one or more of a multi-medium filter, an activated carbon adsorption device, a softener, and a precision filter; if the pre-treatment water treatment system comprises multiple pre-treatment water treatment systems, corresponding parameter adjustment weights are configured for the water treatment performance parameters of the pre-treatment water treatment systems corresponding to the future period according to the process distances between the pre-treatment water treatment systems and the RO system; the parameter adjustment weights are negatively correlated with the process distances.

6. The method according to claim 5, characterized in that, The water inlet characteristics of the multi-medium filter include suspended solids and colloidal particle content; the system working configuration parameters include filter material type, filtration speed, filter core utilization rate, and backwashing frequency and intensity; and the water treatment performance parameters include suspended solids removal rate and turbidity reduction rate. The water inlet characteristics of the activated carbon adsorption device include residual chlorine and organic matter content; the system working configuration parameters include activated carbon type, filling amount, contact time, filter core utilization rate, and water flow speed; and the water treatment performance parameters include residual chlorine removal rate and total organic carbon removal rate. The water inlet characteristics of the softener include calcium and magnesium ion concentration; the system working configuration parameters include ion exchange resin type, ion exchange resin regeneration period, and salt consumption; and the water treatment performance parameters include hardness removal rate and outlet water hardness value. The water inlet characteristics of the precision filter include micro-particle content; the system working configuration parameters include filter core precision, flow, filter core utilization rate, and replacement period; and the water treatment performance parameters include particle interception rate and outlet water turbidity.

7. A control device for an ultrapure water production system, characterized by comprising: The prediction module is configured to input time series data corresponding to a current regulation time into a first target prediction model to generate scale rate information of a future period corresponding to the current regulation time. The time series data includes water inlet parameter characteristics of an RO system in a historical period corresponding to the current regulation time, system working configuration parameter sequences of the RO system in the future period, and water treatment performance parameters of a preposed water treatment system in a preposed future period; the preposed water treatment system is a water treatment system located before the RO system in a process flow sequence; The preposed future period is a period in which the preposed water treatment system has an impact on the RO system in the future period; The recovery rate generation module is configured to generate a recovery rate regulation range of the RO system corresponding to the current regulation time according to the scale rate information of the future period corresponding to the current regulation time. The regulation control module is configured to control a real-time recovery rate of the RO system to dynamically change within the recovery rate regulation range in a regulation period corresponding to the current regulation time. The prediction module is further configured to obtain the water treatment performance parameters of the preposed water treatment system in the future period by the following steps: obtain a water inlet characteristic sequence of the preposed water treatment system in a historical period corresponding to the current regulation time; obtain system working configuration parameter sequences corresponding to the historical period corresponding to the current regulation time and / or the preposed future period, respectively; input the water inlet characteristic sequence and the system working configuration parameter sequence of the preposed water treatment system corresponding to the current regulation time into a second target prediction model to generate the water treatment performance parameters of the preposed water treatment system in the preposed future period; the first target prediction model and the second target prediction model are both LSTM models. The computer program is executed by the processor to implement the control method of the ultrapure water preparation system according to any one of claims 1 to 6. 8.A non-transitory computer-readable storage medium storing a computer program, the computer program comprising instructions configured to cause a processor to perform the method according to any one of claims 1 to 7. The processor executes the computer program to implement the control method of the ultrapure water preparation system according to any one of claims 1 to 6.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, ​

Images