Desalting membrane construction method, device and system

Abstract

The invention discloses a desalting membrane construction method, device and system. The method comprises the following steps: obtaining desalination demand information which is used for representing an expected permeation effect corresponding to a target desalination membrane to be constructed in a mine water concentration application scene; the desalination demand information is input to a target regression analysis model for prediction, target monomer data is determined, the target regression analysis model is obtained by adopting a training data set through machine learning training, and the training data set is constructed based on an academic database corresponding to a mine water concentration application scene; and performing interfacial polymerization treatment according to the target monomer data to construct the target desalination membrane. The technical problems of low efficiency of a desalination membrane construction method and poor performance of the constructed desalination membrane in the prior art are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of desalination membrane construction, and in particular to a desalination membrane construction method, device and system. Background Art

[0002] Mine water is a kind of groundwater or surface water produced in the mining process. The treatment and reuse of mine water is of great significance to environmental protection and resource utilization. Since mine water usually contains high concentrations of salt, heavy metals, suspended matter and organic pollutants, it is necessary to use desalination membranes to moderately desalinate mine water.

[0003] In the related art, desalination membranes are usually set to have high flux and high salt rejection rate. Since mine water does not need excessive desalination, high flux and high salt rejection rate desalination membranes are not suitable for mine water, and new desalination membranes need to be designed to desalinate mine water. However, in the process of designing desalination membranes suitable for mine water, it is necessary to rely on a large number of experiments on mine water manually to obtain the influence of interfacial polymerization monomers of desalination membranes with different chemical structures on the membrane permeability selectivity. The design process is complicated and cumbersome, and is easily affected by human factors, resulting in low design efficiency of mine water desalination membranes, poor performance of the designed desalination membranes, and difficulty in moderately desalting mine water.

[0004] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0005] The embodiments of the present invention provide a desalination membrane construction method, device and system to at least solve the technical problems of low efficiency of the desalination membrane construction method and poor performance of the constructed desalination membrane in the related art.

[0006] According to one aspect of an embodiment of the present invention, a desalination membrane construction method is provided, comprising: obtaining desalination demand information, wherein the desalination demand information is used to characterize an expected permeation effect corresponding to a target desalination membrane to be constructed in a mine water concentration application scenario; inputting the desalination demand information into a target regression analysis model for prediction, and determining target monomer data, wherein the target regression analysis model is obtained through machine learning training using a training data set, and the training data set is constructed based on an academic database corresponding to the mine water concentration application scenario; performing interfacial polymerization treatment according to the target monomer data, and constructing a target desalination membrane.

[0007] Optionally, the desalination membrane construction method further includes: using keywords and academic databases to establish a local data set, wherein the keywords are used to characterize the research topic of constructing a desalination membrane based on interfacial polymerization, and the local data set is used to record the correlation between the key parameters of the monomers corresponding to the desalination membrane and the permeability coefficient; characterizing the local data set to construct a training data set; and using the training data set to train an initial regression analysis model to obtain a target regression analysis model.

[0008] Optionally, the key monomer parameters include: the type of monomer compound, and the proportion of the monomer compound in the desalination membrane.

[0009] Optionally, the local data set is characterized and the construction of the training data set includes: extracting two-dimensional structural features of the local data set to obtain feature descriptors; calling a library function in the target virtual environment to convert the feature descriptors into Morgan fingerprints, wherein the Morgan fingerprints are used to characterize various chemical substructures corresponding to monomer compounds in the local data set; and performing frequency calculations on various chemical substructures corresponding to the Morgan fingerprints to construct the training data set.

[0010] Optionally, the training data set is used to train the initial regression analysis model to obtain the target regression analysis model, including: dividing the training data set using a cross-validation strategy to obtain a training subset and a test subset; constructing the initial regression analysis model based on the initial regression learning algorithm and the initial parameter configuration; using the training subset, performing multiple rounds of iterative training on the initial regression analysis model to obtain an intermediate regression analysis model; using the test subset and preset performance evaluation conditions, performing model verification on the intermediate regression analysis model to obtain the target regression analysis model, wherein the target regression analysis model includes the target regression learning algorithm and target parameter configuration after training adjustment.

[0011] Optionally, the desalination membrane construction method also includes: based on the target regression learning algorithm and the target parameter configuration, performing predicted contribution calculations on the chemical substructures of multiple monomer compounds in the local data set to obtain contribution parameter values; based on the contribution parameter values, constructing multiple chemical substructure libraries, wherein the multiple chemical substructure libraries are used to distinguish the various influence trends of different types of monomer compounds on the permeability coefficient.

[0012] Optionally, the target monomer data includes: at least one monomer compound to be used, and a proportion control equation corresponding to the at least one monomer compound; the desalination demand information is input into a target regression analysis model for prediction, and determining the target monomer data includes: inputting the desalination demand information into a target regression analysis model to screen and determine at least one monomer compound from a plurality of monomer compounds, and performing regression prediction based on the at least one monomer compound to determine the proportion control equation.

[0013] Optionally, the desalination demand information includes an expected range of permeability coefficients, which include water flux and salt rejection.

[0014] According to another aspect of an embodiment of the present invention, a desalination membrane construction device is also provided, including: an acquisition module, used to obtain desalination demand information, wherein the desalination demand information is used to characterize the expected permeation effect corresponding to the target desalination membrane to be constructed in the mine water concentration application scenario; a prediction module, used to input the desalination demand information into a target regression analysis model for prediction, and determine target monomer data, wherein the target regression analysis model is obtained through machine learning training using a training data set, and the training data set is constructed based on an authoritative academic database corresponding to the mine water concentration application scenario; a construction module, used to perform interfacial polymerization treatment according to the target monomer data to construct a target desalination membrane.

[0015] According to another aspect of an embodiment of the present application, a desalination membrane construction system is also provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute any one of the above-mentioned wind power output prediction methods.

[0016] In an embodiment of the present invention, desalination demand information is obtained, wherein the desalination demand information is used to characterize the expected permeability effect corresponding to the target desalination membrane to be constructed in the mine water concentration application scenario; the desalination demand information is input into the target regression analysis model for prediction, and the target monomer data is determined, wherein the target regression analysis model is obtained by machine learning training using a training data set, and the training data set is constructed based on an academic database corresponding to the mine water concentration application scenario; interfacial polymerization treatment is performed according to the target monomer data to construct a target desalination membrane. As a result, the present invention achieves the purpose of constructing a desalination membrane with higher performance more quickly, thereby achieving the technical effect of improving the efficiency of the desalination membrane construction method and improving the performance of the constructed desalination membrane, thereby solving the technical problems of low efficiency of the desalination membrane construction method and poor performance of the constructed desalination membrane in the related art. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0018] Figure 1 is a hardware structure block diagram of an optional terminal device for a desalination membrane construction method according to an embodiment of the present invention;

[0019] Figure 2 is a flow chart of a method for constructing a desalination membrane according to an embodiment of the present invention;

[0020] Figure 3 is a schematic diagram of an optional method for screening monomer compounds of a desalination membrane according to an embodiment of the present invention;

[0021] Figure 4It is a structural block diagram of a desalination membrane construction device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0022] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0023] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0024] According to an embodiment of the present invention, an embodiment of a desalination membrane construction method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0025] Figure 1 is a hardware structure block diagram of a terminal device for a desalination membrane construction method according to an embodiment of the present invention, such as Figure 1As shown, the terminal device may include one or more processors 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor (Microcontroller Unit, referred to as MCU) or a programmable logic device (Field Programmable Gate Array, referred to as FPGA)), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may also include: a display device 110, an input / output device 108 (i.e., an I / O device), a Universal Serial Bus (USB) port (which may be included as one of the ports of a computer bus, not shown in the figure), a network interface (not shown in the figure), a power supply (not shown in the figure) and / or a camera (not shown in the figure). It can be understood by those skilled in the art that Figure 1 The structure shown is for illustration only and does not limit the structure of the above terminal device. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations are shown.

[0026] It should be noted that the one or more processors 102 and / or other data processing circuits may be embodied in whole or in part as software, hardware, firmware or any other combination thereof. In addition, the data processing circuit may be a single independent processing module, or may be fully or partially integrated into any of the other components in the terminal device (or mobile device).

[0027] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the desalination membrane construction method in the embodiment of the present invention. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, realizing the above-mentioned desalination membrane construction method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the terminal device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0028] The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the terminal device. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (Radio Frequency, referred to as RF) module, which is used to communicate with the Internet wirelessly.

[0029] Under the above operating environment, the embodiment of the present invention provides the following Figure 2 The desalination membrane construction method shown in FIG. Figure 2 is a flow chart of a method for constructing a desalination membrane according to an embodiment of the present invention. Figure 2 As shown, the method includes the following implementation steps:

[0030] Step S201, obtaining desalination demand information, wherein the desalination demand information is used to characterize the expected permeability effect corresponding to the target desalination membrane to be constructed in the mine water concentration application scenario;

[0031] Step S202, inputting the desalination demand information into a target regression analysis model for prediction to determine target monomer data, wherein the target regression analysis model is obtained by machine learning training using a training data set, and the training data set is constructed based on an academic database corresponding to a mine water concentration application scenario;

[0032] Step S203, performing interfacial polymerization processing according to the target monomer data to construct a target desalination membrane.

[0033] The above-mentioned desalination demand information may be indicators such as the expected desalination efficiency, the water quality standard after desalination, or the desalination effect under specific conditions. The above-mentioned expected permeation effect may be a performance indicator (such as water flux, desalination rate) that the target desalination membrane needs to achieve, and the expected permeation effect may be used to characterize the specific requirements in the mine water concentration process (such as desalination speed, purity of water after desalination, etc.). The above-mentioned water flux may be the amount of water passing through a unit membrane area per unit time. The above-mentioned desalination rate may be the interception rate of salt in mine water.

[0034] Specifically, experimental data in an academic database corresponding to the mine water concentration application scenario are obtained, for example, experimental data of preparing membranes using different monomer data and performance indicators (such as permeability, desalination rate, etc.) corresponding to the prepared membranes. The experimental data in the academic database are used as a training data set using a machine learning algorithm, and are input into a machine learning algorithm model so that the machine learning algorithm model learns the relationship between the monomer data and the performance indicators corresponding to the desalination membrane, and obtains a target regression analysis model. The above-mentioned machine learning algorithm may include but is not limited to: a linear regression algorithm, a support vector regression algorithm, a decision tree regression algorithm, and a Gaussian process regression algorithm. Further, the performance indicators that the target desalination membrane needs to achieve in the mine water concentration application scenario are collected; using the trained target regression analysis model, the performance indicators that the target desalination membrane needs to achieve are collected as input data, and are input into the target regression analysis model, and the target regression analysis model outputs the target monomer data corresponding to the performance indicators that the target desalination membrane needs to achieve. Furthermore, using the target monomer data, interfacial polymerization monomers (such as water-soluble monomers, oil-soluble monomers) are selected, and interfacial polymerization conditions (such as temperature, time, pH value, solvent selection, etc.) are set to perform interfacial polymerization treatment and construct a target desalination membrane.

[0035] It is easy to understand that through the above steps S201 to S203, in the embodiment of the present application, the academic database and machine learning technology corresponding to the mine water concentration application scenario can be used to more efficiently and accurately screen the desalination membrane monomer data suitable for mine water concentration application. Further, combined with the interfacial polymerization membrane preparation technology, a desalination membrane more suitable for mine water concentration application scenario can be prepared. Therefore, the desalination membrane construction method of the present application has higher construction efficiency and the constructed desalination membrane has better performance.

[0036] The above method of the embodiment of the present invention is further introduced below.

[0037] Optionally, the above desalination membrane construction method may further include the following method steps:

[0038] Step S204, using keywords and academic databases to establish a local data set, wherein the keywords are used to characterize the research topic of building a desalination membrane based on interfacial polymerization, and the local data set is used to record the correlation between the key parameters of the monomers corresponding to the desalination membrane and the permeability coefficient;

[0039] Step S205, performing characterization processing on the local data set to construct a training data set;

[0040] Step S206: Use the training data set to train the initial regression analysis model to obtain a target regression analysis model.

[0041] The above keywords can be words closely related to the research topic of constructing desalination membrane based on interfacial polymerization technology, for example, interfacial polymerization, desalination membrane, permeability coefficient, monomer parameters, construction of separation layer, etc. The keyword can be used to guide literature retrieval and data screening. The keyword can include one word or multiple words. By using keywords, it can be ensured that the collected data is more targeted and relevant. The above permeability coefficient can be a physical parameter corresponding to the ability of fluid to pass through the medium in the desalination membrane. The permeability coefficient can be used to represent the permeability performance of the desalination membrane to the medium. The permeability parameter can include the pure water permeability coefficient and the solute permeability coefficient.

[0042] The above initial regression analysis model may include but is not limited to: a linear regression model, a support vector regression model, a decision tree regression model, and a Gaussian process regression model. The initial regression analysis model may be selected according to actual needs (such as the characteristics of the data set, the expected prediction accuracy, etc.).

[0043] It should be noted that the above local data set can be updated according to actual conditions.

[0044] Specifically, the selected keywords are used to search in the academic database to screen out the relevant research literature on the construction of desalination membranes based on interfacial polymerization technology. The key parameters of the monomers of the desalination membrane (such as monomer type, monomer composition ratio, monomer concentration, monomer reaction temperature, etc.), the permeability coefficient corresponding to the desalination membrane (such as pure water permeation parameter, solute permeation parameter, etc.) and the correlation between the key parameters of the monomers corresponding to the desalination membrane and the permeability coefficient are extracted from the screened research literature, and the correlation between the key parameters of the monomers corresponding to the desalination membrane and the permeability coefficient is recorded in the local data set. Further, the local data set is characterized, and the data in the local data set is converted into a format acceptable to the machine learning model to construct a training data set. Further, the training data set is input into the initial regression analysis model, so that the initial regression analysis model learns the correlation between the key parameters of the monomers corresponding to the desalination membrane and the permeability coefficient, and obtains the target regression analysis model.

[0045] It should be noted that in the process of constructing the local data set, in order to ensure the quality and integrity of the recorded data, the data in the screened research literature can be processed to eliminate noise and outliers in the local data set.

[0046] In an exemplary application scenario, Figure 3 FIG. 1 is a schematic diagram of an optional method for screening interfacial polymerization monomers for desalination membranes according to an embodiment of the present invention. Figure 3As shown, the above academic database can be a reverse osmosis membrane open source database, set keywords, use keywords to search and organize data in the reverse osmosis membrane open source database, and obtain a local data set, for example, a local data set containing the correlation between the molecular structural formulas of 30 reverse osmosis membrane interface polymerization monomers and their corresponding permeability coefficients. Further, the local data set is characterized to obtain a training data set, for example, the correlation between the molecular structural formulas of 30 reverse osmosis membrane interface polymerization monomers in the local data set and their corresponding permeability coefficients is converted into a corresponding string format representing the chemical structure (Simplified Molecular Input Line Entry System, referred to as SMILES). Further, a regression algorithm is selected, and the parameters of the initial regression analysis model are initially configured to construct an initial regression analysis model, and the obtained training data set is input into the initial regression analysis model, and the initial regression analysis model is trained and the parameters of the model are optimized, so that the initial regression analysis model learns the correlation between the key parameters of the monomers corresponding to the desalination membrane and the permeability coefficient, so as to complete the construction of the target regression analysis model.

[0047] It is easy to understand that through the above steps S204 to S206, in the embodiment of the present application, not only can the research results in the existing academic database be fully utilized to more accurately construct a desalination membrane suitable for mine water concentration application scenarios, but also the influence of key monomer parameters on the permeation effect can be automatically identified and learned through machine learning algorithms, so as to more quickly determine the design scheme of the desalination membrane and improve the construction efficiency of the desalination membrane construction method.

[0048] Optionally, in the above desalination membrane construction method, the key parameters of the monomers include: the type of the monomer compound, and the proportion of the monomer compound in the desalination membrane.

[0049] It should be noted that, according to the demand for desalination membranes in different mine water concentration application scenarios, the desalination membrane construction method of the present application can be used to determine the type of corresponding monomer compounds and the proportion of the monomer compounds in the desalination membrane.

[0050] In an exemplary application scenario, Figure 3As shown, the correlation between the key parameters of the monomers corresponding to the above desalination membrane and the permeability coefficient may include: the correlation between the type of monomer compound and the pure water permeability coefficient, the correlation between the type of monomer compound and the solute permeability coefficient, the correlation between the proportion of the monomer compound in the desalination membrane and the pure water permeability coefficient, and the correlation between the proportion of the monomer compound in the desalination membrane and the solute permeability coefficient. The database is searched for data using keywords, and related research on the construction of separation layers based on interfacial polymerization is searched to obtain search results, such as the correlation between the type of monomer compound and the pure water permeability coefficient, the correlation between the type of monomer compound and the solute permeability coefficient, the correlation between the proportion of the monomer compound in the desalination membrane and the pure water permeability coefficient, and the correlation between the proportion of the monomer compound in the desalination membrane and the solute permeability coefficient. The search results are sorted to obtain a local data set.

[0051] It is easy to understand that through the above-mentioned desalination membrane construction method, in the embodiment of the present application, by using the machine learning method, not only the type of monomer compound can be determined, but also the proportion of the monomer compound in the desalination membrane can be determined, and the required structure and performance of the desalination membrane can be determined more accurately. In addition, in the embodiment of the present application, the structure and performance of the desalination membrane can be adjusted according to the demand for desalination membrane in different mine water treatment scenarios, and the corresponding structure and performance of the desalination membrane can be obtained, thereby improving the adaptability of the desalination membrane construction method.

[0052] Optionally, in the above step S205, the local data set is characterized and the construction of the training data set may further include the following execution steps:

[0053] Step S251, extracting two-dimensional structural features from the local data set to obtain a feature descriptor;

[0054] Step S252, calling a library function in the target virtual environment to convert the feature descriptor into a Morgan fingerprint, wherein the Morgan fingerprint is used to characterize a variety of chemical substructures corresponding to monomer compounds in the local data set;

[0055] Step S253, performing frequency calculation on the various chemical substructures corresponding to the Morgan fingerprints to construct a training data set.

[0056] The above two-dimensional structural features may be structural features of the compound without considering its three-dimensional spatial configuration.

[0057] The above-mentioned feature descriptors can be numerical information corresponding to the structural information of the monomer compound, and the feature descriptors can include but are not limited to atomic types, bond types, frequency of occurrence of functional groups, size and number of rings, molecular weight, number of hydrogen bond donors and acceptors, etc.

[0058] The above-mentioned Morgan fingerprint can be a coding method for describing the structure of a compound.

[0059] Specifically, chemical informatics tools (such as RDKit, OpenBabel, etc.) are used to read the structural information of monomer compounds recorded in the local data set, perform two-dimensional structural feature extraction, and convert the structural information of monomer compounds into a series of numerical feature descriptors. Furthermore, by calling the library function in the target virtual environment, the chemical structure characteristics of the compound and the influence of the local environment on the atoms are analyzed, and the feature descriptors are converted into the Morgan fingerprints of the corresponding multiple chemical substructures. Each chemical substructure in the corresponding multiple chemical substructures in the Morgan fingerprint is statistically analyzed to obtain the frequency calculation results of each chemical substructure. The frequency calculation results of each chemical substructure in the Morgan fingerprint are also used as the characteristics of the chemical substructure to construct a training data set.

[0060] In an exemplary application scenario, Figure 3 As shown, the above-mentioned characterization processing can be Morgan fingerprint processing and frequency calculation processing, and the above-mentioned training data set can be a data set of chemical substructures of interfacial polymerization monomers. The above-mentioned target virtual environment can be an integrated development environment (IDE) of Python, for example, a Pycharm environment. The correlation between the molecular structural formulas of 30 kinds of reverse osmosis membrane interfacial polymerization monomers in the local data set and their corresponding permeability coefficients is extracted by two-dimensional structural features, and the two-dimensional structures of 30 kinds of reverse osmosis membrane interfacial polymerization monomers are converted into feature descriptors (such as SMILES strings) that can be recognized by the machine learning model, and the data set corresponding to the converted SMILES strings is loaded into the Pycharm environment, and the feature descriptors are converted into Morgan fingerprints by calling library functions, and the corresponding x kinds of chemical substructures are generated, and the frequency of occurrence of each chemical substructure is calculated and sorted, and the top y (y≤x) chemical substructures are screened out, and the multiple screened chemical substructures are standardized and the abnormal data are cleaned up, so as to construct a data set of chemical substructures of interfacial polymerization monomers.

[0061] It is easy to understand that, through the above steps S251 to S253, the embodiment of the present application performs two-dimensional structural feature extraction on the local data set to obtain feature descriptors, which can simplify the data processing process, effectively improve the efficiency and ability of the computer to process data, and at the same time enhance the interpretability and operability of the data set. By performing Morgan fingerprint processing and frequency calculation processing on the feature descriptors, the chemical properties of the monomer compounds can be more comprehensively reflected, the quality of the training data set can be improved, and thus the accuracy of the target regression analysis model can be improved, ensuring that the constructed desalination membrane can achieve the expected desalination effect.

[0062] Optionally, in the above step S206, using the training data set to train the initial regression analysis model to obtain the target regression analysis model may also include the following execution steps:

[0063] Step S261, using a cross-validation strategy to divide the training data set into a training subset and a test subset;

[0064] Step S262, constructing an initial regression analysis model based on the initial regression learning algorithm and the initial parameter configuration;

[0065] Step S263, using the training subset, performing multiple rounds of iterative training on the initial regression analysis model to obtain an intermediate regression analysis model;

[0066] Step S264, using the test subset and the preset performance evaluation conditions, the intermediate regression analysis model is verified to obtain the target regression analysis model, wherein the target regression analysis model includes the trained and adjusted target regression learning algorithm and target parameter configuration.

[0067] The above cross-validation strategy can be a statistical method for evaluating model performance. The cross-validation strategy can ensure the generalization ability of the model by dividing the training data set into multiple subsets. The cross-validation strategy can include but is not limited to: K-Fold Cross-validation, Leave-One-Out Cross-validation (LOOCV for short), Time Series Cross-validation, and Rolling Cross-validation.

[0068] The above-mentioned training subset can be used to train the model. The above-mentioned test subset can be used to verify and evaluate the model. The above-mentioned initial regression learning algorithm may include but is not limited to: Bayesian Ridge regression algorithm, linear regression algorithm, elastic network algorithm, support vector regression algorithm, gradient boosting regression algorithm and random forest algorithm, and the initial regression learning algorithm may include at least one regression learning algorithm. The above-mentioned initial parameters may include but are not limited to: the depth of the decision tree, the number of layers of the neural network, the learning rate, etc. The above-mentioned initial parameter configuration can be preliminarily set based on experience. The above-mentioned preset performance evaluation conditions can be used to evaluate the performance of the model, and the preset performance evaluation conditions may include but are not limited to the mean square error index, the mean absolute error index, etc. By using the preset performance evaluation conditions, the performance of different regression analysis models can be compared. The above-mentioned intermediate regression analysis model can be a regression analysis model obtained by training the initial regression analysis model using the training subset.

[0069] Specifically, a cross-validation strategy is used to divide the training data set into a training subset and a test subset. Further, at least one initial regression learning algorithm is selected, and based on each initial regression learning algorithm, the initial parameters corresponding to the initial regression learning algorithm are configured to construct an initial regression analysis model corresponding to the initial regression learning algorithm, thereby obtaining at least one initial regression analysis model. Further, using the training subset, multiple rounds of iterative training are performed on each initial regression analysis model. During the training process of the initial regression analysis model, the parameters of the model can be adjusted by using a parameter optimization method (such as grid search, etc.) to minimize the loss function, thereby obtaining a corresponding intermediate regression analysis model, thereby obtaining at least one intermediate regression analysis model. The performance of each intermediate regression analysis model is verified using the test subset to obtain a verification result, and the performance of each intermediate regression analysis model is analyzed using a preset performance evaluation condition to obtain an analysis result. Based on the verification result and the analysis result, a target regression analysis model is selected from the intermediate regression analysis model to obtain a target regression learning algorithm and target parameter configuration after training adjustment corresponding to the target regression analysis model.

[0070] In an exemplary application scenario, Figure 3As shown, the training data set can be a chemical substructure data set about interfacial polymerization monomers. The cross-validation strategy can be a K-fold cross-validation method. The initial regression analysis algorithm can include a Bayesian ridge regression algorithm, a linear regression algorithm, an elastic network algorithm, a support vector regression algorithm, a gradient boosting regression algorithm, and a random forest algorithm. The chemical substructure data set about interfacial polymerization monomers is divided using the K-fold cross-validation method to obtain a training subset and a test subset, and the initial parameters corresponding to each initial regression analysis algorithm are configured (e.g., the initial parameters corresponding to the Bayesian ridge regression algorithm are configured), and the initial regression analysis model corresponding to the initial regression learning algorithm is constructed (e.g., an initial Bayesian ridge regression analysis model), and accordingly, an initial linear regression analysis model, an initial elastic network analysis model, an initial support vector regression analysis model, an initial gradient boosting regression analysis model, and an initial random forest analysis model can also be obtained. Use the training subset to perform multiple rounds of iterative training on each initial regression analysis model to obtain an intermediate regression analysis model (such as an intermediate Bayesian Ridge regression analysis model). Correspondingly, an intermediate linear regression analysis model, an intermediate elastic network analysis model, an intermediate support vector regression analysis model, an intermediate gradient boosting regression analysis model, and an intermediate random forest analysis model can also be obtained. Further, use the test subset to verify each intermediate regression analysis model to obtain the verification result corresponding to each intermediate regression analysis model, use the preset performance evaluation conditions to analyze the performance of each intermediate regression analysis model to obtain the analysis result corresponding to each intermediate regression analysis model, and determine the target regression analysis model based on the verification results and analysis results corresponding to multiple intermediate regression analysis models, and obtain the target regression learning algorithm (such as the gradient boosting regression analysis model) and the target parameter configuration.

[0071] It is easy to understand that through the above steps S261 to S264, on the one hand, in the embodiment of the present application, through cross-validation and iterative training, it is possible to effectively avoid model overfitting, ensure the generalization ability of the model on different data sets, and be suitable for the construction and optimization of desalination membranes for large-scale mine water treatment projects. On the other hand, in the embodiment of the present application, by constructing at least one initial regression analysis model, using a training subset, a test subset, and preset performance evaluation conditions, a variety of regression analysis models can be obtained, thereby selecting a regression learning algorithm and parameter configuration corresponding to the target desalination membrane required for the mine water concentration application scenario, which can improve the performance of the constructed desalination membrane.

[0072] Optionally, the above desalination membrane construction method may further include the following method steps:

[0073] Step S207, based on the target regression learning algorithm and the target parameter configuration, the predicted contribution calculation is performed on the chemical substructures of the multiple monomer compounds in the local data set to obtain the contribution parameter value;

[0074] Step S208: construct multiple chemical substructure libraries according to the contribution parameter values, wherein the multiple chemical substructure libraries are used to distinguish the various influence trends of different types of monomer compounds on the permeability coefficient.

[0075] The predicted contribution value can be used to indicate the contribution degree of the chemical substructure feature of each monomer compound to the prediction of the permeability coefficient. The contribution parameter value can be a quantitative value of the contribution degree of the chemical substructure feature of each monomer compound to the permeability coefficient.

[0076] Specifically, the predicted contribution of the chemical substructure of each of the multiple monomer compounds in the local data set is calculated using the determined target regression learning algorithm and target parameter configuration to obtain a contribution parameter value. Based on the contribution parameter values ​​of the chemical substructures, multiple chemical substructure libraries can be constructed, and the chemical substructure libraries contain multiple chemical substructures that have multiple effects on the permeability coefficient.

[0077] It should be noted that in the process of constructing multiple chemical substructure libraries, at least one contribution threshold can be set. Based on the set contribution threshold and the contribution parameter value of the chemical substructure, the chemical substructures of multiple monomer compounds are analyzed, and the chemical substructures of the multiple monomer compounds are divided into different chemical substructure libraries to construct multiple chemical substructure libraries.

[0078] In an exemplary application scenario, Figure 3 As shown, the target regression learning algorithm is a gradient boosting regression algorithm, and the target parameter can be the fitting degree of the correlation between the monomer compound concentration and the membrane permeability coefficient performance of the desalination membrane, and the target parameter can be configured as 0.99. Based on the gradient boosting regression algorithm and the fitting degree of the correlation between the monomer compound concentration and the membrane permeability coefficient performance of the desalination membrane, the chemical substructures of multiple monomer compounds in the local data set are predicted to calculate the contribution degree, and the contribution parameter values ​​corresponding to the chemical substructures of multiple monomer compounds are obtained. Further, it is determined whether the contribution parameter value corresponding to the chemical substructure of each monomer compound in the multiple monomer compounds is a positive value. If the contribution parameter value corresponding to the chemical substructure of the monomer compound is a positive value, the chemical substructure of the monomer compound is classified into the chemical substructure library of water permeability positive effect; if the contribution parameter value corresponding to the chemical substructure of the monomer compound is not a positive value, the chemical substructure of the monomer compound is classified into the chemical substructure library of desalination rate negative effect.

[0079] It is easy to understand that, through the above steps S207 to S208, in the embodiment of the present application, the contribution of the chemical substructure of each monomer compound to the permeability coefficient can be quantified, and the influence of the chemical substructure of each monomer compound on the desalination membrane performance can be more accurately determined, so that a desalination membrane that is more suitable for mine water concentration application scenarios can be designed.

[0080] Optionally, the target monomer data includes: at least one monomer compound to be used, and a proportional control equation corresponding to the at least one monomer compound; in the above step S202, the desalination demand information is input into the target regression analysis model for prediction, and determining the target monomer data may also include the following execution steps:

[0081] Step S211 , inputting the desalination demand information into the target regression analysis model to screen and determine at least one monomer compound from a plurality of monomer compounds, and performing regression prediction based on the at least one monomer compound to determine a proportional control equation.

[0082] The above-mentioned proportion control equation can be a mathematical expression for designing composite materials. The proportion control equation can be used to express the relationship between the relative proportions of different monomer compounds in the mixture and the desired performance. The proportion control equation can be constructed through a machine learning model.

[0083] Specifically, the desalination demand information is input as input data into the target regression analysis model. The target regression analysis model calculates the prediction result of each monomer compound among the multiple monomer compounds under a given desalination demand based on the relationship between the structural characteristics of the monomer compounds and the performance of the constructed desalination membrane learned during the training process, and obtains multiple prediction results. At least one monomer compound that meets the desalination demand information is screened out from the multiple prediction results, and the target regression analysis model is used to perform regression prediction on the at least one monomer compound, so as to determine the corresponding proportional control equation to improve the performance of the constructed desalination membrane.

[0084] It is easy to understand that through the above step S211, in the embodiment of the present application, the monomer compounds that meet the specific desalination requirements can be quickly screened out through the machine learning model, and the proportional control equation can be predicted based on the chemical substructure characteristics of these compounds, thereby improving the efficiency and accuracy of the desalination membrane construction method.

[0085] Optionally, in the above desalination membrane construction method, the desalination demand information includes an expected range of permeability coefficient, and the permeability coefficient includes water flux and salt rejection.

[0086] It is easy to understand that, through the above desalination membrane construction method, in the embodiments of the present application, by setting the expected permeability coefficient range, it can be ensured that the constructed desalination membrane has a good salt rejection rate while ensuring high water flux, thereby effectively improving the performance of the constructed desalination membrane.

[0087] In this embodiment, a desalination membrane construction device is also provided, which is used to implement the above-mentioned embodiments and preferred embodiments, and the descriptions that have been made will not be repeated. As used below, a "module" is a combination of software and / or hardware that can implement a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0088] Figure 4 is a structural block diagram of a desalination membrane construction device according to an embodiment of the present invention. Figure 4 As shown, the device includes: an acquisition module 401, used to obtain desalination demand information, wherein the desalination demand information is used to characterize the expected permeability effect corresponding to the target desalination membrane to be constructed in the mine water concentration application scenario; a prediction module 402, used to input the desalination demand information into a target regression analysis model for prediction, and determine target monomer data, wherein the target regression analysis model is obtained by machine learning training using a training data set, and the training data set is constructed based on an authoritative academic database corresponding to the mine water concentration application scenario; a construction module 403, used to perform interfacial polymerization according to the target monomer data to construct the target desalination membrane.

[0089] Optionally, in addition to all the above modules, the above-mentioned desalination membrane construction device also includes a training module (not shown in the figure), which is used to: use keywords and academic databases to establish a local data set, wherein the keywords are used to characterize the research topic of constructing desalination membranes based on interfacial polymerization, and the local data set is used to record the correlation between the key parameters of the monomers corresponding to the desalination membrane and the permeability coefficient; characterize the local data set to construct a training data set, and use the training data set to train the initial regression analysis model to obtain the target regression analysis model.

[0090] Optionally, the key monomer parameters include: the type of monomer compound, and the proportion of the monomer compound in the desalination membrane.

[0091] Optionally, the above training module can also be used to: extract two-dimensional structural features of the local data set to obtain feature descriptors; call the library function in the target virtual environment to convert the feature descriptors into Morgan fingerprints, wherein the Morgan fingerprints are used to characterize the various chemical substructures corresponding to the monomer compounds in the local data set; perform frequency calculations on the various chemical substructures corresponding to the Morgan fingerprints to construct a training data set.

[0092] Optionally, the above-mentioned training module can also be used to: divide the training data set using a cross-validation strategy to obtain a training subset and a test subset; construct an initial regression analysis model based on an initial regression learning algorithm and an initial parameter configuration; use the training subset to perform multiple rounds of iterative training on the initial regression analysis model to obtain an intermediate regression analysis model; use the test subset and preset performance evaluation conditions to perform model verification on the intermediate regression analysis model to obtain a target regression analysis model, wherein the target regression analysis model includes the target regression learning algorithm and target parameter configuration after training adjustment.

[0093] Optionally, in addition to all the above modules, the above-mentioned desalination membrane construction device also includes a construction module (not shown in the figure), which is used to: based on the target regression learning algorithm and the target parameter configuration, calculate the predicted contribution of the chemical substructures of multiple monomer compounds in the local data set to obtain contribution parameter values; according to the contribution parameter values, construct multiple chemical substructure libraries, wherein the multiple chemical substructure libraries are used to distinguish the various influence trends of different types of monomer compounds on the permeability coefficient.

[0094] Optionally, the target monomer data includes: at least one monomer compound to be used, and a proportional control equation corresponding to the at least one monomer compound; the above-mentioned prediction module 402 can also be used to: input the desalination demand information into the target regression analysis model to screen and determine at least one monomer compound from a plurality of monomer compounds, and perform regression prediction based on the at least one monomer compound to determine the proportional control equation.

[0095] Optionally, the desalination demand information includes an expected range of permeability coefficients, which include water flux and salt rejection.

[0096] It should be noted that the above modules can be implemented by software or hardware. For the latter, it can be implemented in the following ways, but not limited to: the above modules are all located in the same processor; or the above modules are located in different processors in any combination.

[0097] Through the device technical solutions corresponding to the above-mentioned acquisition module, prediction module, and construction module, the embodiment of the present application obtains desalination demand information through the acquisition module, wherein the desalination demand information is used to characterize the expected permeability effect corresponding to the target desalination membrane to be constructed in the mine water concentration application scenario; through the prediction module, the desalination demand information is input into the target regression analysis model for prediction to determine the target monomer data, wherein the target regression analysis model is obtained through machine learning training using a training data set, and the training data set is constructed based on an authoritative academic database corresponding to the mine water concentration application scenario; through the construction module, interfacial polymerization treatment is performed according to the target monomer data to construct the target desalination membrane.

[0098] In the above-mentioned wind power output prediction device provided by the embodiment of the present invention, desalination demand information is obtained, wherein the desalination demand information is used to characterize the expected permeability effect corresponding to the target desalination membrane to be constructed in the mine water concentration application scenario; the desalination demand information is input into the target regression analysis model for prediction, and the target monomer data is determined, wherein the target regression analysis model is obtained by machine learning training using a training data set, and the training data set is constructed based on an academic database corresponding to the mine water concentration application scenario; interfacial polymerization treatment is performed according to the target monomer data to construct a target desalination membrane. As a result, the present invention achieves the purpose of constructing a desalination membrane with higher performance more quickly, thereby achieving the technical effect of improving the efficiency of the desalination membrane construction method and improving the performance of the constructed desalination membrane, thereby solving the technical problems of low efficiency of the desalination membrane construction method and poor performance of the constructed desalination membrane in the related technology.

[0099] According to another aspect of an embodiment of the present invention, a desalination membrane construction system is provided, comprising a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute any one of the above-mentioned desalination membrane construction methods.

[0100] Optionally, in this embodiment, the processor can be configured to perform the following steps through a computer program: obtaining desalination demand information, wherein the desalination demand information is used to characterize the expected permeation effect corresponding to the target desalination membrane to be constructed in the mine water concentration application scenario; inputting the desalination demand information into a target regression analysis model for prediction, and determining target monomer data, wherein the target regression analysis model is obtained through machine learning training using a training data set, and the training data set is constructed based on an academic database corresponding to the mine water concentration application scenario; performing interfacial polymerization processing according to the target monomer data to construct a target desalination membrane.

[0101] Optionally, the specific examples in this embodiment may refer to the examples described in the above embodiment and its optional implementation manners, which will not be described in detail here.

[0102] The serial numbers of the above embodiments of the present invention are for description only and do not represent the advantages or disadvantages of the embodiments.

[0103] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0104] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0105] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0106] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0107] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, ROM, RAM, mobile hard disks, magnetic disks or optical disks.

[0108] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for constructing a desalination membrane, characterized in that: include: Obtaining desalination demand information, wherein the desalination demand information is used to characterize the expected permeability effect corresponding to the target desalination membrane to be constructed in the mine water concentration application scenario; The desalination demand information is input into a target regression analysis model for prediction to determine target monomer data, wherein the target regression analysis model is obtained through machine learning training using a training data set, and the training data set is constructed based on an academic database corresponding to the mine water concentration application scenario; An interfacial polymerization process is performed according to the target monomer data to construct the target desalination membrane.

2. The method for constructing a desalination membrane according to claim 1, characterized in that: The desalination membrane construction method also includes: Using the keywords and the academic database, a local data set is established, wherein the keywords are used to characterize the research topic of constructing a desalination membrane based on interfacial polymerization, and the local data set is used to record the correlation between the key parameters of the monomers corresponding to the desalination membrane and the permeability coefficient; Performing characterization processing on the local data set to construct the training data set; The training data set is used to train the initial regression analysis model to obtain the target regression analysis model.

3. The method for constructing a desalination membrane according to claim 2, characterized in that: The key parameters of the monomer include: the type of the monomer compound and the proportion of the monomer compound in the desalination membrane.

4. The method for constructing a desalination membrane according to claim 2, characterized in that: Performing characterization processing on the local data set to construct the training data set includes: Extracting two-dimensional structural features from the local data set to obtain a feature descriptor; Calling a library function in the target virtual environment to convert the feature descriptor into a Morgan fingerprint, wherein the Morgan fingerprint is used to characterize a plurality of chemical substructures corresponding to the monomer compound in the local data set; The frequency of the multiple chemical substructures corresponding to the Morgan fingerprints is calculated to construct the training data set.

5. The method for constructing a desalination membrane according to claim 2, characterized in that: Using the training data set to train the initial regression analysis model to obtain the target regression analysis model includes: The training data set is divided into a training subset and a test subset by adopting a cross-validation strategy; Constructing the initial regression analysis model based on the initial regression learning algorithm and the initial parameter configuration; Using the training subset, performing multiple rounds of iterative training on the initial regression analysis model to obtain an intermediate regression analysis model; The intermediate regression analysis model is validated by using the test subset and preset performance evaluation conditions to obtain the target regression analysis model, wherein the target regression analysis model includes the trained and adjusted target regression learning algorithm and target parameter configuration.

6. The method for constructing a desalination membrane according to claim 5, characterized in that: The desalination membrane construction method also includes: Based on the target regression learning algorithm and the target parameter configuration, performing prediction contribution calculation on the chemical substructures of the plurality of monomer compounds in the local data set to obtain contribution parameter values; According to the contribution parameter value, a plurality of chemical substructure libraries are constructed, wherein the plurality of chemical substructure libraries are used to distinguish various influence trends of different types of monomer compounds on the permeability coefficient.

7. The method for constructing a desalination membrane according to claim 6, characterized in that: The target monomer data includes: at least one monomer compound to be used, and a proportional control equation corresponding to the at least one monomer compound; the desalination demand information is input into a target regression analysis model for prediction, and the target monomer data is determined to include: The desalination demand information is input into the target regression analysis model to screen and determine the at least one monomer compound from the plurality of monomer compounds, and the proportional control equation is determined by regression prediction based on the at least one monomer compound.

8. The method for constructing a desalination membrane according to any one of claims 1 to 7, characterized in that: The desalination demand information includes an expected range of permeability coefficients, which include water flux and salt rejection.

9. A desalination membrane construction device, characterized in that: include: An acquisition module, used for acquiring desalination demand information, wherein the desalination demand information is used to characterize the expected permeability effect corresponding to the target desalination membrane to be constructed in the mine water concentration application scenario; A prediction module, used for inputting the desalination demand information into a target regression analysis model for prediction and determining target monomer data, wherein the target regression analysis model is obtained by machine learning training using a training data set, and the training data set is constructed based on an authoritative academic database corresponding to the mine water concentration application scenario; A construction module is used to perform interfacial polymerization according to the target monomer data to construct the target desalination membrane.

10. A desalination membrane construction system, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute the desalination membrane construction method according to any one of claims 1 to 8.

Images