A method for optimizing the production of PES microporous membranes
Through sampling detection and cast membrane partition search, the phase conversion parameters are optimized, and the parameter control problem in PES microporous filter membrane production is solved, and the performance stability of the filter membrane is improved.
Patent Information
- Application Number
- CN202411062904.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-05
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2044-08-05
AI Technical Summary
It is difficult to accurately control the phase conversion parameters in PES microporous filter membrane production, resulting in unstable filter membrane performance.
The solution uniformity information and component distribution information were obtained through sampling detection, and the cast film partition search and phase conversion parameter optimization were performed, and the cast film partition search direction was adjusted until the search optimization of all cast film partitions was completed to obtain the optimal phase conversion parameter set.
Accurate control and optimization of phase conversion parameters is achieved, and the performance stability of PES microporous filter membrane is improved.
Smart Images

Figure CN118577143B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to data processing related fields, and in particular to a method for optimizing the production of a PES microporous filter membrane. Background Art
[0002] PES microporous membrane is a filter membrane with a microporous structure made of polyethersulfone (PES) resin, which is used to filter and separate particles and solutes in liquids. In the production process of PES microporous membrane, phase inversion is a crucial step, which determines the key properties of the membrane, such as pore size, distribution and filtration efficiency. In the traditional phase inversion process, parameter control is usually based on experience and trial and error. Due to the complexity of the phase inversion process, the interaction and influence between some parameters are difficult to predict and control, resulting in poor instability and repeatability of the membrane production results, and it is difficult to achieve precise control of membrane performance.
[0003] In the current related technologies, the production of PES microporous filter membranes has technical problems such as difficulty in achieving precise control and optimization of phase conversion parameters and unstable filter membrane performance. Summary of the invention
[0004] The present application provides a method for optimizing the production of PES microporous filter membranes, which adopts technical means such as sampling detection, obtaining solution uniformity information and component distribution information, searching for casting membrane partitions, and optimizing phase transformation parameters of casting membrane partitions, thereby achieving the technical effect of realizing precise control and optimization of phase transformation parameters and improving the performance stability of PES microporous filter membranes.
[0005] The present application provides a method for optimizing the production of a PES microporous filter membrane, comprising:
[0006] A polymer solution formed by mixing polyethersulfone resin and solvent is sampled and tested to obtain solution uniformity information and component distribution information, wherein the polymer solution is used to produce a PES microporous filter membrane, and the component distribution information includes multiple component information; the polymer solution is used to perform a casting process to obtain a cast film; according to the solution uniformity information, a detection step length for detecting the cast film is analyzed; using the detection step length, a casting film partition search is performed within the cast film, and the searched casting film partition information is indexed within the component distribution information to obtain indexed component information; according to the indexed component information, the phase transformation parameters of the casting film partition are optimized to obtain the optimal phase transformation parameters; according to the casting film partition information, the casting film partition search direction is adjusted, and the search and phase transformation parameter optimization of other casting film partitions are continued until the search and optimization of all casting film partitions are completed to obtain the optimal phase transformation parameter set, and all casting film partitions are subjected to phase transformation processing and production to obtain a PES microporous filter membrane.
[0007] In a possible implementation, a polymer solution formed by mixing a polyethersulfone resin and a solvent is sampled and tested to obtain solution uniformity information and component distribution information, and the following processing is performed:
[0008] A polymer solution formed by mixing a polyethersulfone resin and a solvent is sampled to obtain a plurality of sample solutions; the plurality of sample solutions are tested to obtain a plurality of component information; solution uniformity information is calculated based on the plurality of component information; the plurality of component information is clustered, and the component distribution information is constructed based on the number of component information in each clustering result.
[0009] In a possible implementation, according to the solution uniformity information, the detection step length for detecting the cast film is analyzed, and the following processing is performed:
[0010] According to the production data records of the PES microporous filter membrane, a set of sample solution uniformity information is collected, and the minimum step length for uneven changes in the cast membrane under different sample solution uniformity information is collected to obtain a sample detection step length set; a detection step length decision maker is constructed using the sample solution uniformity information set and the sample detection step length set; based on the detection step length decision maker, a decision is made on the solution uniformity information to obtain the detection step length.
[0011] In a possible implementation, the detection step length is used to search for a casting film partition in the casting film, the searched casting film partition information is indexed in the component distribution information to obtain the index component information, and the following processing is performed:
[0012] The cast film is subjected to coordinate identification processing to obtain the cast film position space; a plurality of first search partitions are randomly selected in the cast film position space according to the size of the detection step; a plurality of first cast film partition information of the plurality of first search partitions are collected, and indexed in the component distribution information to obtain a plurality of first index component information.
[0013] In a possible implementation, a plurality of first casting film partition information of a plurality of first search partitions is collected, indexed in the component distribution information, a plurality of first index component information is obtained, and the following processing is performed:
[0014] Collect the first cast film partition images of multiple first search partitions; identify the multiple first cast film partition images to obtain multiple first cast film partition information, wherein a cast film information identifier is trained using sample cast film partition images and sample cast film partition information to identify the cast film partition images; obtain the sample cast film partition information set, and obtain the sample solution component information set, and construct a mapping relationship; based on the multiple first cast film partition information, index mapping obtains multiple first index component information in the component distribution information.
[0015] In a possible implementation, the phase transformation parameters of the casting film partition are optimized according to the index component information to obtain the optimal phase transformation parameters, and the following processing is performed: a phase transformation function for optimizing the phase transformation parameters of the casting film partition is constructed, as shown in the following formula: ; Among them, PTF is the phase transformation fitness, and are the particle size weight and uniformity weight, The predicted filtration particle size information is obtained by using the optimized phase transformation parameters combined with the first index component information to predict the phase transformation. is the preset filtering particle size information, K is the particle size uniformity information obtained by using the optimized phase transformation parameters combined with the first index component information to predict the phase transformation; the phase transformation parameter space is obtained, and a plurality of first phase transformation parameters are randomly generated in the phase transformation parameter space; according to the plurality of first phase transformation parameters, in combination with the plurality of first index component information, phase transformation prediction is performed to obtain a plurality of first predicted filtering particle size information and a plurality of first particle size uniformity information, and a plurality of first phase transformation fitnesses are calculated and obtained; the phase transformation parameters of the plurality of first search partitions are continuously optimized until convergence, and a plurality of phase transformation parameters with the largest phase transformation fitness are output to obtain a plurality of optimal phase transformation parameters.
[0016] In a possible implementation, the phase transformation prediction is performed according to the plurality of first phase transformation parameters and the plurality of first index component information, and the following processing is performed:
[0017] Collect a set of sample solution component information, and based on the casting film phase transformation data record, collect a set of sample phase transformation parameters, a set of sample filtration particle size information, and a set of sample particle size uniformity information; use the sample solution component information set, the sample phase transformation parameter set, the sample filtration particle size information set, and the sample particle size uniformity information set to train a phase transformation predictor; use the phase transformation predictor to perform phase transformation prediction on the multiple first phase transformation parameters combined with the multiple first index component information to obtain multiple first predicted filtration particle size information and multiple first particle size uniformity information.
[0018] In a possible implementation, according to the casting film partition information, the casting film partition search direction is adjusted, and the search and phase transformation parameter optimization are continued for other casting film partitions until the search and optimization of all casting film partitions are completed, and the following processing is performed:
[0019] According to multiple random search directions, a search is performed based on the first search partition to obtain multiple second search partitions; multiple second casting film partition information of the multiple second search partitions are collected; it is determined whether the multiple second casting film partition information are the same as the multiple first casting film partition information, if so, the search direction is not adjusted, and the optimal phase transformation parameters in the first search partition are used as the optimal phase transformation parameters in the second search partition, if not, the search direction is randomly adjusted; the remaining casting film partitions are continued to be searched and the phase transformation parameter optimization is optimized until the search optimization of all casting film partitions is completed.
[0020] A PES microporous filter membrane production optimization method proposed in the present application is to first sample and detect a polymer solution formed by mixing a polyethersulfone resin and a solvent to obtain solution uniformity information and component distribution information, wherein the polymer solution is used to produce a PES microporous filter membrane, and the component distribution information includes multiple component information. Then, the polymer solution is used for film casting to obtain a cast film, and then the detection step length for detecting the cast film is analyzed according to the solution uniformity information, and then the detection step length is used to search for the cast film partitions in the cast film, and the searched cast film partition information is indexed in the component distribution information to obtain the indexed component information, and then the phase transformation parameters of the cast film partitions are optimized according to the indexed component information to obtain the optimal phase transformation parameters, and finally, according to the cast film partition information, the search direction of the cast film partition is adjusted, and the search and phase transformation parameter optimization of other cast film partitions are continued until the search and optimization of all cast film partitions are completed, and the optimal phase transformation parameter set is obtained, and all cast film partitions are subjected to phase transformation production to obtain PES microporous filter membrane, thereby achieving the technical effect of realizing accurate control and optimization of phase transformation parameters and improving the performance stability of PES microporous filter membrane. BRIEF DESCRIPTION OF THE DRAWINGS
[0021] In order to more clearly illustrate the technical solution of the embodiment of the present invention, the accompanying drawings of the embodiment of the present invention will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the method according to the embodiment of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.
[0022] Figure 1 A schematic diagram of a process flow of a PES microporous filter membrane production optimization method provided in an embodiment of the present application.
[0023] Figure 2 A schematic diagram of a process for obtaining multiple first index component information in a PES microporous filter membrane production optimization method provided in an embodiment of the present application. DETAILED DESCRIPTION
[0024] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.
[0025] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.
[0026] In the following description, reference is made to “some embodiments”, which describe a subset of all possible embodiments, but it is understood that “some embodiments” may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms “first\second” involved are merely to distinguish similar objects and do not represent a specific ordering of the objects. The terms “including” and “having” and any variations are intended to cover non-exclusive inclusions. For example, a process, method, product, or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by technicians in the technical field to which this application belongs. The terms used herein are for the purpose of describing the embodiments of this application only. The embodiments of this application provide a method for optimizing the production of a PES microporous filter membrane, such as Figure 1 As shown, the method includes:
[0027] Step S100, sampling and testing the polymer solution formed by mixing polyethersulfone resin and solvent to obtain solution uniformity information and component distribution information, wherein the polymer solution is used to produce PES microporous filter membrane, and the component distribution information includes multiple component information. Specifically, according to a predetermined formula, the polyethersulfone resin and the solvent are mixed in a certain proportion, and the solution is fully stirred by a stirring tool until the resin is completely dissolved and the solution is in a uniform state, thereby obtaining a polymer solution. In the polymer solution, a sampling tool is used to randomly select samples from multiple different positions for extraction, and a transparency meter, a microscope, a laser particle size analyzer and other equipment are used to perform uniformity analysis on the samples, including analyzing the distribution of particles in the polymer solution, particle size, etc.; a chromatograph, a mass spectrometer or a spectrum analyzer and other equipment are used to perform component analysis on the extracted samples to determine the content of polyethersulfone resin in the polymer solution, the type and content of the solvent and other key component information. All data obtained from the sampling test are recorded to obtain solution uniformity information and component distribution information.
[0028] In a possible implementation, step S100 further includes step S110, sampling a polymer solution formed by mixing polyethersulfone resin and solvent to obtain multiple sample solutions. Specifically, using a sampling tool, such as a sampling tube, a sampler, etc., randomly select different positions in the polymer solution for sampling, and extract a sufficient number of sample solutions. Step S120, detecting the multiple sample solutions to obtain multiple component information. Specifically, using a chemical analysis instrument (such as a chromatograph, a mass spectrometer, etc.) to detect the sample solution, detect the content of polyethersulfone resin, solvent and other additives or impurities, and obtain the component information of each sample solution, including the concentration or content of each component. Step S130, according to the multiple component information, calculate and obtain the solution uniformity information. Specifically, compare the component information of multiple sample solutions, analyze the distribution of each component in different samples, use statistical methods (such as variance, standard deviation, etc.) to calculate the fluctuation of the concentration of each component, and evaluate the uniformity of the solution according to the fluctuation. The smaller the fluctuation, the better the uniformity. Step S140, clustering the multiple component information, and constructing the component distribution information according to the number of component information in each clustering result. Specifically, clustering algorithms (such as K-means, hierarchical clustering, etc.) are used to cluster the component information of multiple sample solutions, and according to the clustering results, samples with similar component information are classified into the same category, and the number and distribution of samples in each category are analyzed to construct component distribution information, which includes information such as the location, size, and component content of different component areas. This implementation method uses sampling and cluster analysis methods to conduct a detailed and accurate assessment of the uniformity and component distribution of the polymer solution, achieving the technical effect of comprehensively and accurately assessing the uniformity and component distribution of the polymer solution.
[0029] Step S200, using the polymer solution to perform film casting treatment to obtain a cast film. Specifically, the polymer solution is transferred to a liquid storage tank of a film casting machine (equipment for coating the polymer solution on a carrier and controlling solvent volatilization and film layer solidification), the film casting machine is started, and the polymer solution is evenly coated on a film casting carrier (a substrate for carrying the polymer solution and forming a film layer, such as a glass plate, a metal mesh, etc.) through a scraper or a spinneret (forming a film layer by spraying the polymer solution).
[0030] Step S300, analyzing the detection step length for detecting the cast film according to the solution uniformity information. Specifically, analyzing the influence of solution uniformity on the quality of the cast film. The uniformity of the solution directly affects the structure and performance of the cast film. If the uniformity of the solution is good (i.e., the fluctuation is small), the cast film can show higher uniformity and fewer defects. According to the cast film production process and quality control requirements, the corresponding relationship between the detection step length and the solution uniformity is set in combination with historical data, industry standards or expert suggestions, wherein the detection step length refers to the distance or interval that the detection equipment moves along the surface of the cast film during the cast film detection process, which is used to determine the degree of detail of the detection. For example, a threshold is set. When the solution uniformity is lower than the threshold, a smaller detection step length is used; otherwise, a larger detection step length is used. According to the solution uniformity information and the set corresponding relationship, the detection step length during the cast film detection is calculated. According to the calculated detection step length, the parameters of the cast film detection equipment, such as the scanning speed, the sampling frequency, etc., are adjusted so that the detection equipment can accurately detect the cast film with the set detection step length.
[0031] In a possible implementation, step S300 further includes step S310, collecting a sample solution uniformity information set according to the production data record of the PES microporous filter membrane, and collecting the minimum step length of the casting film under different sample solution uniformity information to obtain a sample detection step length set. Specifically, the production data record of the PES microporous filter membrane is retrieved from the production log, that is, various data generated during the production process of the PES microporous filter membrane, including raw material information, process parameters, test results, etc., and the parts related to solution uniformity and casting film quality are collected from the production data record. Multiple sample solution uniformity information is extracted from these records to form a set. For each sample solution uniformity information, the corresponding casting film detection data is searched to determine at which step length the casting film begins to change unevenly, and the minimum step length of these uneven changes is recorded to form a sample detection step length set. Step S320, using the sample solution uniformity information set and the sample detection step length set, a detection step length decision maker is constructed. Specifically, machine learning, regression analysis or other statistical methods are used to model the sample solution uniformity information set and the sample detection step set, and a decision maker that can predict the appropriate detection step according to the solution uniformity information is constructed. The detection step decision maker is trained and verified to ensure the accuracy and reliability of its prediction results. Step S330, based on the detection step decision maker, a decision is made on the solution uniformity information to obtain the detection step. Specifically, the currently obtained solution uniformity information is input into the detection step decision maker, and the detection step decision maker predicts a suitable detection step according to the input solution uniformity information. This implementation method accurately analyzes the relationship between solution uniformity and cast film quality by collecting and analyzing a large amount of sample data, and realizes automatic adjustment of the detection step by using the detection step decision maker for prediction, thereby achieving the technical effect of improving detection efficiency and accuracy.
[0032] Step S400, using the detection step length, searching for the casting film partitions in the casting film, indexing the searched casting film partition information in the component distribution information, and obtaining the indexed component information. Specifically, according to the detection step length, the casting film is divided into equal-sized partitions. For example, if the casting film is two-dimensional, these partitions may be rectangular or square; if it is three-dimensional, they may be cubes or other shapes. Starting from a starting point of the casting film, each casting film partition is searched in turn according to a predetermined path (such as snake, zigzag, etc.), random order or other order, and for each casting film partition, an imaging device or a measuring instrument is used to collect its image or related physical parameters (such as uniformity, thickness, density, etc.). Based on historical production data or experimental data, an index system between polymer solution components and casting film partition characteristics is established. This index system can be based on a database, lookup table or other data structure. By comparing the relationship between the casting film partition characteristics (such as uniformity, thickness, etc.) and the polymer solution components recorded in the index system, the collected casting film partition information is matched with the data in the index system. Once a match is found, the corresponding solution composition information, i.e., the index component information, is extracted from the index system.
[0033] In a possible implementation, step S400 further includes step S410, performing coordinate identification processing on the cast film to obtain the cast film position space. Specifically, determine the size and shape of the cast film, set a two-dimensional or three-dimensional coordinate system for the cast film, so that each point on the cast film has a unique coordinate identification, which is used to locate and identify different positions on the cast film, and incorporate all points of the cast film into the coordinate system to form the cast film position space, that is, the cast film position space is a set of coordinates of all points on the cast film, representing the position and shape of the cast film in a specific coordinate system. Step S420, randomly select multiple first search partitions in the cast film position space according to the size of the detection step. Specifically, in the cast film position space, according to the size of the detection step, randomly select multiple starting points, and define multiple first search partitions with these starting points as the center according to the size of the detection step. Step S430, collect multiple first cast film partition information of multiple first search partitions, index them in the component distribution information, and obtain multiple first index component information. Specifically, the first cast film partition information of each first search partition is collected and acquired by using imaging equipment, measuring equipment, etc. An index mechanism is constructed based on historical data, experimental results or expert knowledge. The index mechanism can be a mapping relationship or database, which contains information about the cast film partition characteristics and the corresponding solution components, and can associate the cast film partition information with the component distribution information. For example, by analyzing the characteristics of the cast film partitions under different solution components, a lookup table or model can be established to map the cast film partition information to the corresponding component information. The first cast film partition information is input into the index mechanism for matching, and the index mechanism outputs multiple first index component information corresponding to each first cast film partition information. This implementation method adopts coordinate identification processing and random selection of search partitions. The coordinate identification processing provides accurate positioning and identification capabilities, and the random selection of search partitions ensures the comprehensiveness of the search, thereby achieving the technical effect of improving the accuracy and efficiency of the analysis.
[0034] like Figure 2As shown, in a possible implementation, step S430 further includes step S431, collecting first cast film partition images of multiple first search partitions. Specifically, use an imaging device (such as an optical microscope, a scanning electron microscope, etc.) to image each first search partition. Step S432, identify the multiple first cast film partition images to obtain multiple first cast film partition information, wherein the cast film information identifier is trained using the sample cast film partition image and the sample cast film partition information to identify the cast film partition image. Specifically, the captured first cast film partition image is identified using a trained cast film information identifier (a tool or algorithm for identifying cast film partition information), and the cast film information identifier extracts the features of the first cast film partition image by comparing the sample cast film partition image and the sample cast film partition information. Based on these features, the cast film information identifier generates the corresponding first cast film partition information (such as uniformity, thickness, etc.). Step S433, obtain the sample cast film partition information set, and obtain the sample solution component information set, and construct a mapping relationship. Specifically, known sample casting film partition information is collected to form a sample casting film partition information set, sample solution component information corresponding to these sample casting film partition information is obtained to form a sample solution component information set, and a mapping relationship between the sample casting film partition information set and the sample solution component information set is established using statistical or machine learning techniques. Step S434, based on multiple first casting film partition information, index mapping obtains multiple first index component information in the component distribution information. Specifically, the identified multiple first casting film partition information is input into the constructed mapping relationship, and the mapping relationship outputs the corresponding multiple first index component information, which reflects the solution components corresponding to the first search partition. This implementation method accurately infers the solution components corresponding to the search partition through the mapping relationship constructed by the sample casting film partition information and the sample solution component information, and achieves the technical effect of accurately obtaining the relationship between the properties of different regions in the casting film and the solution components.
[0035] Step S500, according to the index component information, the phase transformation parameters of the casting film partition are optimized to obtain the optimal phase transformation parameters. Specifically, the index component information includes the specific components, concentrations and other key parameters of the solution in the casting film partition. The index component information is analyzed to determine the influence mechanism of different components on the phase transformation process, for example, how different components affect the uniformity of the filter membrane after phase transformation and the accuracy of the filter particle size. According to the properties of the casting film material and the index component information, the key parameters in the phase transformation process are set based on historical practical experience, theoretical models or industry standards, that is, various parameters that affect the performance and structure of the filter membrane during the phase transformation process, such as temperature, time, humidity composition, etc. A mathematical model between the phase transformation parameters and the filter membrane performance (uniformity of the filter membrane, accuracy of the filter particle size, etc.) is established using statistical analysis methods, regression analysis or machine learning algorithms, and based on the established mathematical model, the phase transformation parameters are optimized to find the optimal parameter combination that can achieve the best performance of the filter membrane, that is, the optimal phase transformation parameters.
[0036] In a possible implementation, step S500 further includes step S510, constructing a phase transformation function for optimizing the phase transformation parameters of the casting film partition, as shown in the following formula:
[0037] ; Among them, PTF is the phase transformation fitness, and are the particle size weight and uniformity weight, The predicted filtration particle size information is obtained by using the optimized phase transformation parameters combined with the first index component information to predict the phase transformation. is the preset filter particle size information, K is the particle size uniformity information obtained by using the optimized phase transformation parameters combined with the first index component information to predict the phase transformation. Specifically, the phase transformation fitness is a quantitative indicator used to evaluate the performance of the filter membrane under different phase transformation parameter combinations, the particle size weight indicates the importance of the filter particle size in the optimization target, the uniformity weight indicates the importance of the filter particle size uniformity in the optimization target, the predicted filter particle size information is the predicted value of the filter membrane filter particle size obtained by phase transformation prediction, the preset filter particle size information is the filter particle size value expected to be achieved by the filter membrane, and the particle size uniformity information indicates the uniformity of the particle size distribution in the filter membrane. Step S520, obtain the phase transformation parameter space, and randomly generate multiple first phase transformation parameters in the phase transformation parameter space. Specifically, according to the properties of the casting material, the preparation process and other factors, the possible value range of the phase transformation parameter is determined to form a phase transformation parameter space, and a plurality of first phase transformation parameter combinations are randomly generated in the phase transformation parameter space. Step S530, according to the multiple first phase transformation parameters, combined with the multiple first index component information, phase transformation prediction is performed to obtain multiple first predicted filtered particle size information and multiple first particle size uniformity information, and multiple first phase transformation fitness is calculated. Specifically, each generated first phase transformation parameter combination is used in combination with the corresponding first index component information to perform phase transformation prediction through experiments, simulations or empirical formulas, and the predicted filtered particle size information and particle size uniformity information are extracted from the prediction results. The prediction results are substituted into the phase transformation function, and the phase transformation fitness corresponding to each phase transformation parameter combination is calculated. Step S540, the phase transformation parameters of the multiple first search partitions are continuously optimized until convergence, and multiple phase transformation parameters with the largest phase transformation fitness are output to obtain multiple optimal phase transformation parameters. Specifically, according to the calculated phase transformation fitness, an optimization algorithm (such as a genetic algorithm, a particle swarm algorithm, etc.) is used to optimize the phase transformation parameters of multiple first search partitions, and steps S520 and S530 are repeated to continuously update the phase transformation parameter combination until the convergence condition is met (such as the fitness is no longer significantly improved, the preset number of iterations is reached, etc.), and the phase transformation parameter combination with the largest phase transformation fitness is selected as the optimal phase transformation parameter output. This implementation method constructs a phase transformation function and combines the random generation and optimization methods to quickly find the phase transformation parameter combination that optimizes the performance of the filter membrane in the phase transformation parameter space, thereby achieving the technical effect of improving the optimization efficiency.
[0038] In a possible implementation, according to the multiple first phase transformation parameters, combined with the multiple first index component information, phase transformation prediction is performed, and step S530 further includes step S531, collecting a sample solution component information set, and collecting a sample phase transformation parameter set, a sample filtration particle size information set and a sample particle size uniformity information set according to the casting film phase transformation data record. Specifically, a representative sample solution is selected, and the sample solution has different components and concentrations. A chemical analysis instrument (such as a chromatograph, a mass spectrometer, etc.) is used to measure the content of each component in the sample solution and record it as a sample solution component information set; according to the known casting film phase transformation process, key parameters of the sample in the phase transformation process, such as temperature, humidity, pH value, etc., are obtained or measured, and recorded as a sample phase transformation parameter set; a particle size analyzer and other equipment are used to measure the filtration particle size of the sample after phase transformation, and recorded as a sample filtration particle size information set; the uniformity of the sample particle size is evaluated by statistical or other methods, and recorded as a sample particle size uniformity information set. Step S532, using the sample solution component information set, sample phase transformation parameter set, sample filtration particle size information set and sample particle size uniformity information set to train a phase transformation predictor. Specifically, the collected sample set is cleaned, normalized and other preprocessing operations are performed, and a suitable prediction model, such as a neural network, a decision tree, a support vector machine, etc., is selected or designed, and the preprocessed sample solution component information set and the sample phase transformation parameter set are used as inputs, and the corresponding sample filtration particle size information set and the sample particle size uniformity information set are used as outputs to train the prediction model to obtain a phase transformation predictor. Step S533, using the phase transformation predictor, the multiple first phase transformation parameters are combined with the multiple first index component information to perform phase transformation prediction, and multiple first predicted filtration particle size information and multiple first particle size uniformity information are obtained. Specifically, the multiple first phase transformation parameters and the multiple first index component information are organized into a prediction data set, and the prediction data is input into the trained phase transformation predictor to obtain the corresponding predicted filtration particle size information and particle size uniformity information. This implementation method collects and analyzes sample data, uses machine learning or statistical learning methods to train a phase transformation predictor, and quickly and accurately predicts the filtration particle size and particle size uniformity after phase transformation, achieving the technical effect of improving the efficiency and accuracy of phase transformation prediction.
[0039] Step S600, according to the cast film partition information, the search direction of the cast film partition is adjusted, and the search and phase transformation parameter optimization of other cast film partitions are continued until the search and optimization of all cast film partitions are completed, and the optimal phase transformation parameter set is obtained, and all cast film partitions are subjected to phase transformation processing and production to obtain PES microporous filter membrane. Specifically, after completing the search and phase transformation parameter optimization of one or more cast film partitions, the subsequent cast film partition search direction is adjusted according to the obtained cast film partition information and the optimal phase transformation parameters, and the cast film partitions with the same or similar performance are preferentially searched and optimized. According to the adjusted search direction, other cast film partitions are continued to be searched, and for each newly searched cast film partition, steps S400 and S500 are repeated to optimize the phase transformation parameters until all cast film partitions are searched and optimized, and a set containing the optimal phase transformation parameters of all cast film partitions is obtained. The obtained optimal phase transformation parameter set is applied to the actual casting film production process. According to the optimal phase transformation parameters of each casting film partition, the corresponding production equipment and process parameters are adjusted, and all casting film partitions are subjected to phase transformation production. By controlling parameters such as temperature, time, and coagulation bath composition, the casting film is phase separated to form a filter membrane with a microporous structure, thereby obtaining the PES microporous filter membrane. The embodiment of the present application adopts technical means such as sampling detection, obtaining solution uniformity information and component distribution information, searching for casting film partitions, and optimizing the phase transformation parameters of casting film partitions, so as to achieve the technical effect of realizing accurate control and optimization of phase transformation parameters and improving the performance stability of the PES microporous filter membrane.
[0040] In a possible implementation, according to the casting film partition information, the casting film partition search direction is adjusted, and the other casting film partitions are continuously searched and the phase transformation parameters are optimized until the search optimization of all casting film partitions is completed. Step S600 further includes step S610, searching based on the first search partition according to multiple random search directions to obtain multiple second search partitions. Specifically, according to a certain strategy (such as uniform distribution, Gaussian distribution, etc.), multiple search directions are randomly generated, that is, directions or paths from one partition to another partition, starting from the first search partition, searching according to the randomly generated search directions to obtain multiple second search partitions. Step S620, collecting multiple second casting film partition information of the multiple second search partitions. Specifically, for each second search partition, collect its related casting film partition information, including partition position, material, uniformity, size, etc., and record the collected casting film partition information as multiple second casting film partition information. Step S630, respectively determine whether the multiple second casting film partition information is the same as the multiple first casting film partition information. If so, the search direction is not adjusted, and the optimal phase transformation parameters in the first search partition are used as the optimal phase transformation parameters in the second search partition. If not, the search direction is randomly adjusted. Specifically, each second casting film partition information is compared with the first casting film partition information before the search. If the second casting film partition information is the same as the first casting film partition information (or similar to a certain threshold), it is considered that the two casting film partitions are similar and have similar phase transformation parameter requirements. There is no need to adjust the search direction, and the optimal phase transformation parameters in the first search partition are directly applied to the second search partition; if the second casting film partition information is different from the first casting film partition information, the search direction is randomly adjusted and the search is performed again. Step S640, continue to search and optimize the phase transformation parameters of the remaining casting film partitions until the search optimization of all casting film partitions is completed. Specifically, steps S610 to S630 are repeated until all the casting film partitions are searched and the phase transformation parameters are optimized. In other words, during the search process, priority is given to casting film partitions that are identical or similar to the casting film partition information that has been searched, and these casting film partitions have similar phase transformation parameter requirements. During the search, the phase transformation parameters of the casting film partitions that are large in area and meet the requirements are optimized first. For some special casting film partitions that are small in area or have special requirements for phase transformation conditions, they can be processed after the optimization of the large-area casting film partitions is completed. This implementation method saves time and computing resources by preferentially searching and processing the same or similar casting film partitions, and achieves the technical effect of improving the efficiency of phase transformation parameter optimization.
[0041] The above specific implementation manner does not constitute a limitation to the protection scope of the present application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the protection scope of the present application. In some cases, the actions or steps recorded in the present application can be performed in an order different from that in the embodiment and can still achieve the desired results. In addition, the process depicted in the accompanying drawings does not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
Claims
1. A method for optimizing the production of PES microporous membranes, characterized in that: The method comprises: Sampling and testing a polymer solution formed by mixing a polyethersulfone resin and a solvent to obtain solution uniformity information and component distribution information, wherein the polymer solution is used to produce a PES microporous filter membrane, and the component distribution information includes multiple component information; Using the polymer solution to perform film casting treatment to obtain a cast film; Analyzing the detection step length for detecting the cast film according to the solution uniformity information; The detection step length is adopted to search for the casting film partitions in the casting film, and the searched casting film partition information is indexed in the component distribution information to obtain the index component information, wherein a plurality of first search partitions are randomly selected, and a plurality of first casting film partition information of the plurality of first search partitions are collected, and indexed in the component distribution information to obtain a plurality of first index component information; according to the index component information, the phase transformation parameters of the casting film partitions are optimized to obtain the optimal phase transformation parameters, wherein a phase transformation function for optimizing the phase transformation parameters of the casting film partitions is constructed, and the phase transformation parameters are optimized, and the phase transformation function is as follows: Among them, PTF is the phase transformation fitness, and are the particle size weight and uniformity weight, The predicted filtration particle size information is obtained by using the optimized phase transformation parameters combined with the first index component information to predict the phase transformation. is the preset filtered particle size information, K is the particle size uniformity information obtained by using the optimized phase transformation parameters combined with the first index component information to predict the phase transformation; According to the casting film partition information, the casting film partition search direction is adjusted, and the other casting film partitions are continuously searched and the phase transformation parameter optimization is performed until the search optimization of all casting film partitions is completed, and the optimal phase transformation parameter set is obtained, and all casting film partitions are subjected to phase transformation processing and production to obtain PES microporous filter membrane; Among them, according to the index component information, the phase transformation parameters of the casting film partition are optimized to obtain the optimal phase transformation parameters, including: Acquiring a phase transformation parameter space, and randomly generating a plurality of first phase transformation parameters in the phase transformation parameter space; According to the plurality of first phase transformation parameters, combined with the plurality of first index component information, phase transformation prediction is performed to obtain a plurality of first predicted filtered particle size information and a plurality of first particle size uniformity information, and a plurality of first phase transformation fitness degrees are calculated; Continue to optimize the phase transformation parameters of the multiple first search partitions until convergence, output multiple phase transformation parameters with the largest phase transformation fitness, and obtain multiple optimal phase transformation parameters; According to the plurality of first phase transformation parameters, combined with the plurality of first index component information, phase transformation prediction is performed, including: Collecting a sample solution component information set, and collecting a sample phase transformation parameter set, a sample filtration particle size information set, and a sample particle size uniformity information set based on the film casting phase transformation data record; The sample solution component information set, the sample phase transformation parameter set, the sample filtration particle size information set and the sample particle size uniformity information set are used to train a phase transformation predictor; the phase transformation predictor is used to perform phase transformation prediction on the multiple first phase transformation parameters in combination with the multiple first index component information to obtain multiple first predicted filtration particle size information and multiple first particle size uniformity information.
2. A PES microporous membrane production optimization method according to claim 1, characterized in that: The polymer solution formed by mixing polyethersulfone resin and solvent is sampled and tested to obtain solution uniformity information and component distribution information, including: Sampling a polymer solution formed by mixing a polyethersulfone resin and a solvent to obtain a plurality of sample solutions; Testing the multiple sample solutions to obtain multiple component information; Calculating and obtaining solution uniformity information according to the plurality of component information; The plurality of component information are clustered, and the component distribution information is constructed and obtained according to the amount of component information in each clustering result.
3. A PES microporous membrane production optimization method according to claim 1, characterized in that: Analyzing the detection step length for detecting the cast film according to the solution uniformity information includes: According to the production data records of the PES microporous filter membrane, a set of sample solution uniformity information is collected, and the minimum step length of uneven changes in the casting membrane under different sample solution uniformity information is collected to obtain a sample detection step length set; Using the sample solution uniformity information set and the sample detection step set, constructing a detection step decision device; Based on the detection step length decider, a decision is made on the solution uniformity information to obtain the detection step length.
4. A PES microporous membrane production optimization method according to claim 1, characterized in that: The detection step length is used to search for a casting film partition in the casting film, and the searched casting film partition information is indexed in the component distribution information to obtain indexed component information, including: Performing coordinate identification processing on the cast film to obtain the cast film position space; According to the size of the detection step length, randomly selecting a plurality of first search partitions in the film casting position space; Collect a plurality of first casting film partition information of a plurality of first search partitions, perform indexing in the component distribution information, and obtain a plurality of first index component information.
5. A PES microporous membrane production optimization method according to claim 4, characterized in that: Collecting a plurality of first casting film partition information of a plurality of first search partitions, indexing in the component distribution information, and obtaining a plurality of first index component information, including: Collecting first cast film subarea images of a plurality of first search subareas; Recognize the plurality of first cast film partition images to obtain a plurality of first cast film partition information, wherein a cast film information identifier is trained using sample cast film partition images and sample cast film partition information to recognize the cast film partition images; Obtaining a sample film casting partition information set and a sample solution component information set, and constructing a mapping relationship; Based on the plurality of first casting film partition information, the index mapping obtains a plurality of first index component information in the component distribution information.
6. A PES microporous membrane production optimization method according to claim 5, characterized in that: According to the casting film partition information, the casting film partition search direction is adjusted, and other casting film partitions are continuously searched and the phase transformation parameter optimization is performed until the search optimization of all casting film partitions is completed, including: Searching based on the first search partition according to multiple random search directions to obtain multiple second search partitions; Collecting a plurality of second casting film partition information of the plurality of second search partitions; Determine respectively whether the multiple second casting film partition information are the same as the multiple first casting film partition information. If so, do not adjust the search direction, and use the optimal phase transformation parameters in the first search partition as the optimal phase transformation parameters in the second search partition. If not, randomly adjust the search direction; continue to search and optimize the phase transformation parameters of the remaining casting film partitions until the search and optimization of all casting film partitions are completed.
