CN114934198B - Method for preparing high-purity indium based on machine learning optimization vacuum distillation - Google Patents
Method for preparing high-purity indium based on machine learning optimization vacuum distillation Download PDFInfo
- Publication number
- CN114934198B CN114934198B CN202210487994.9A CN202210487994A CN114934198B CN 114934198 B CN114934198 B CN 114934198B CN 202210487994 A CN202210487994 A CN 202210487994A CN 114934198 B CN114934198 B CN 114934198B
- Authority
- CN
- China
- Prior art keywords
- vacuum distillation
- indium
- machine learning
- purity
- purity indium
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000005292 vacuum distillation Methods 0.000 title claims abstract description 97
- 229910052738 indium Inorganic materials 0.000 title claims abstract description 89
- APFVFJFRJDLVQX-UHFFFAOYSA-N indium atom Chemical compound [In] APFVFJFRJDLVQX-UHFFFAOYSA-N 0.000 title claims abstract description 89
- 238000000034 method Methods 0.000 title claims abstract description 67
- 238000010801 machine learning Methods 0.000 title claims abstract description 56
- 238000005457 optimization Methods 0.000 title claims abstract description 15
- 238000000746 purification Methods 0.000 claims abstract description 19
- 238000004519 manufacturing process Methods 0.000 claims abstract description 17
- 238000005868 electrolysis reaction Methods 0.000 claims abstract description 13
- 239000002994 raw material Substances 0.000 claims abstract description 4
- 238000004821 distillation Methods 0.000 claims description 21
- 229910052751 metal Inorganic materials 0.000 claims description 14
- 239000002184 metal Substances 0.000 claims description 14
- FAPWRFPIFSIZLT-UHFFFAOYSA-M Sodium chloride Chemical compound [Na+].[Cl-] FAPWRFPIFSIZLT-UHFFFAOYSA-M 0.000 claims description 6
- 239000007791 liquid phase Substances 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 238000001514 detection method Methods 0.000 claims description 5
- 239000003792 electrolyte Substances 0.000 claims description 5
- 239000012535 impurity Substances 0.000 claims description 5
- 238000005070 sampling Methods 0.000 claims description 5
- 238000013528 artificial neural network Methods 0.000 claims description 4
- 238000001035 drying Methods 0.000 claims description 4
- 238000001036 glow-discharge mass spectrometry Methods 0.000 claims description 4
- 238000010438 heat treatment Methods 0.000 claims description 4
- 238000004321 preservation Methods 0.000 claims description 4
- 238000004506 ultrasonic cleaning Methods 0.000 claims description 4
- 238000012795 verification Methods 0.000 claims description 4
- 108010010803 Gelatin Proteins 0.000 claims description 3
- RTAQQCXQSZGOHL-UHFFFAOYSA-N Titanium Chemical compound [Ti] RTAQQCXQSZGOHL-UHFFFAOYSA-N 0.000 claims description 3
- 239000000654 additive Substances 0.000 claims description 3
- 229920000159 gelatin Polymers 0.000 claims description 3
- 239000008273 gelatin Substances 0.000 claims description 3
- 235000019322 gelatine Nutrition 0.000 claims description 3
- 235000011852 gelatine desserts Nutrition 0.000 claims description 3
- 229910000337 indium(III) sulfate Inorganic materials 0.000 claims description 3
- XGCKLPDYTQRDTR-UHFFFAOYSA-H indium(iii) sulfate Chemical compound [In+3].[In+3].[O-]S([O-])(=O)=O.[O-]S([O-])(=O)=O.[O-]S([O-])(=O)=O XGCKLPDYTQRDTR-UHFFFAOYSA-H 0.000 claims description 3
- 239000011780 sodium chloride Substances 0.000 claims description 3
- 239000010936 titanium Substances 0.000 claims description 3
- 229910052719 titanium Inorganic materials 0.000 claims description 3
- 230000000630 rising effect Effects 0.000 claims 1
- 239000004065 semiconductor Substances 0.000 abstract description 6
- 239000007769 metal material Substances 0.000 abstract description 3
- 238000002474 experimental method Methods 0.000 description 5
- 239000000463 material Substances 0.000 description 5
- 150000001875 compounds Chemical class 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000001174 ascending effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000002790 cross-validation Methods 0.000 description 2
- 238000007418 data mining Methods 0.000 description 2
- 238000002360 preparation method Methods 0.000 description 2
- 239000010453 quartz Substances 0.000 description 2
- VYPSYNLAJGMNEJ-UHFFFAOYSA-N silicon dioxide Inorganic materials O=[Si]=O VYPSYNLAJGMNEJ-UHFFFAOYSA-N 0.000 description 2
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 1
- GPXJNWSHGFTCBW-UHFFFAOYSA-N Indium phosphide Chemical compound [In]#P GPXJNWSHGFTCBW-UHFFFAOYSA-N 0.000 description 1
- 229910052785 arsenic Inorganic materials 0.000 description 1
- RQNWIZPPADIBDY-UHFFFAOYSA-N arsenic atom Chemical compound [As] RQNWIZPPADIBDY-UHFFFAOYSA-N 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 229910052793 cadmium Inorganic materials 0.000 description 1
- BDOSMKKIYDKNTQ-UHFFFAOYSA-N cadmium atom Chemical compound [Cd] BDOSMKKIYDKNTQ-UHFFFAOYSA-N 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000001816 cooling Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000013210 evaluation model Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 239000011521 glass Substances 0.000 description 1
- 229910002804 graphite Inorganic materials 0.000 description 1
- 239000010439 graphite Substances 0.000 description 1
- WPYVAWXEWQSOGY-UHFFFAOYSA-N indium antimonide Chemical compound [Sb]#[In] WPYVAWXEWQSOGY-UHFFFAOYSA-N 0.000 description 1
- RPQDHPTXJYYUPQ-UHFFFAOYSA-N indium arsenide Chemical compound [In]#[As] RPQDHPTXJYYUPQ-UHFFFAOYSA-N 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000005086 pumping Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000007711 solidification Methods 0.000 description 1
- 230000008023 solidification Effects 0.000 description 1
- 238000005092 sublimation method Methods 0.000 description 1
- 229910052714 tellurium Inorganic materials 0.000 description 1
- PORWMNRCUJJQNO-UHFFFAOYSA-N tellurium atom Chemical compound [Te] PORWMNRCUJJQNO-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- C—CHEMISTRY; METALLURGY
- C22—METALLURGY; FERROUS OR NON-FERROUS ALLOYS; TREATMENT OF ALLOYS OR NON-FERROUS METALS
- C22B—PRODUCTION AND REFINING OF METALS; PRETREATMENT OF RAW MATERIALS
- C22B58/00—Obtaining gallium or indium
-
- C—CHEMISTRY; METALLURGY
- C22—METALLURGY; FERROUS OR NON-FERROUS ALLOYS; TREATMENT OF ALLOYS OR NON-FERROUS METALS
- C22B—PRODUCTION AND REFINING OF METALS; PRETREATMENT OF RAW MATERIALS
- C22B9/00—General processes of refining or remelting of metals; Apparatus for electroslag or arc remelting of metals
- C22B9/02—Refining by liquating, filtering, centrifuging, distilling, or supersonic wave action including acoustic waves
-
- C—CHEMISTRY; METALLURGY
- C22—METALLURGY; FERROUS OR NON-FERROUS ALLOYS; TREATMENT OF ALLOYS OR NON-FERROUS METALS
- C22B—PRODUCTION AND REFINING OF METALS; PRETREATMENT OF RAW MATERIALS
- C22B9/00—General processes of refining or remelting of metals; Apparatus for electroslag or arc remelting of metals
- C22B9/04—Refining by applying a vacuum
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P10/00—Technologies related to metal processing
- Y02P10/20—Recycling
Landscapes
- Engineering & Computer Science (AREA)
- Chemical & Material Sciences (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Manufacturing & Machinery (AREA)
- Mechanical Engineering (AREA)
- Metallurgy (AREA)
- Organic Chemistry (AREA)
- Software Systems (AREA)
- Materials Engineering (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Acoustics & Sound (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Manufacture And Refinement Of Metals (AREA)
- Vaporization, Distillation, Condensation, Sublimation, And Cold Traps (AREA)
Abstract
A method for preparing high-purity indium based on machine learning optimization vacuum distillation selects 5N indium prepared by electrolysis as a raw material, and performs vacuum distillation and twice purification by a vacuum distillation furnace; according to the technological parameters obtained by the two-time purification, a high-purity indium experimental database is constructed, then a machine learning method is utilized to assist in resolving the vacuum distillation process, a multi-factor coupled machine learning model is established, the product quality is predicted, and experimental technological parameters in a certain range are optimized, so that the aim of optimizing the high-purity indium vacuum distillation purification process is fulfilled. According to the invention, a vacuum distillation optimization model is established by means of a machine learning method, the accuracy of the vacuum distillation model is checked through database data, and the vacuum distillation experimental technological parameters are predicted and optimized, so that the vacuum distillation purification process optimization of the high-purity indium is realized, the purity of the high-purity indium can be improved, the production technological parameters are solidified, the production controllability is enhanced, the production efficiency is improved, and high-quality high-purity metal materials are provided for the semiconductor industry.
Description
Technical Field
The invention belongs to the technical field of nonferrous metal vacuum distillation purification, and particularly relates to a method for preparing high-purity indium based on machine learning optimization vacuum distillation.
Background
The high-purity indium has wide application in the field of photoelectrons and compound semiconductors thereof, and is mainly synthesized into indium-based III-V compound semiconductors such as indium antimonide (InSb), indium phosphide (InP), indium arsenide (InAs) and the like, and the high-purity indium-containing compound semiconductors are used as laser light sources for optical fiber communication, heterojunction solar cell materials, infrared detection and magneto-optical devices, and have higher and higher purity requirements on indium metal, generally 6N and above, and the impurity content is required to be lower than 1ppm and even lower than 0.1ppm. The preparation methods of high-purity indium are numerous, and mainly include sublimation method, directional solidification method, vacuum distillation method, extraction method, electrolytic method and the like. The vacuum distillation method has high efficiency, is easy to operate and can be widely applied to the field of metal purification.
Machine learning theory is mainly the design and analysis of algorithms that allow computers to automatically "learn". The machine learning algorithm is an algorithm for automatically analyzing and identifying patterns and statistical rules from data and predicting unknown data by utilizing the rules. Machine learning has found very wide application, for example: data mining, computer vision, natural language processing, etc. The vacuum distillation purification process is researched, the influence rule of distillation process parameters of the high-purity indium on the purification effect is excavated, and the internal relation between the process parameters and the product performance (purity) is established, so that the method has important significance for realizing the stable mass production of the high-purity indium of 6N and above. By means of a machine learning method, a large number of reliable experiments are combined with data mining, internal relations or experience rules are mined, numerical simulation optimization results are used for guiding the experiments, meanwhile, machine learning is utilized for establishing a model and predicting an optimal test parameter range, optimal test parameters are obtained, further, reverse prediction is conducted for guiding production, the purity of high-purity indium of a product is optimized through machine learning, optimal process parameters are searched, controllability is improved, efficiency is improved, and the method is beneficial to providing high-quality high-purity metal materials for the semiconductor industry.
Disclosure of Invention
The invention aims to provide a method for preparing high-purity indium by optimizing a vacuum distillation process based on a machine learning model, which optimizes the vacuum distillation purification of the high-purity indium by using the machine learning method to obtain a high-purity indium product with higher purity.
The technical scheme adopted by the invention is as follows:
a method for preparing high-purity indium based on machine learning optimization vacuum distillation comprises the following steps:
s1) preparing 5N indium by using an electrolysis method as a raw material: the electrolyte system of the electrolytic method is indium sulfate solution containing additives of sodium chloride and gelatin, the pH value is 1.5-3.0, and the current density is 30-85A/m 2 Under the electrolysis condition that the electrolysis temperature is 20-30 ℃, 3-4N metal indium is melted and cast to be used as an anode, a titanium plate is used as a cathode plate for electrolysis to obtain metal indium, and the metal indium is cleaned and dried to obtain 5N indium;
s2) putting 5N indium into a vacuum distillation furnace for vacuum distillation and twice purification: placing 5N indium into a crucible of a vacuum distillation furnace, vacuumizing, and reducing the vacuum degree to 10 -1 When Pa is lower, the temperature is set to 800-1200 ℃, the primary distillation is carried out for 3-8 h, then the residual liquid phase bottom component is taken to carry out the secondary distillation at 1200-1500 ℃ for 5-8 h, after the heating and heat preservation stage is finished, the heating program and the vacuumizing system are closed, the furnace is cooled to room temperature, the component volatilized from the top in the vacuum distillation is sampled, the ultrasonic cleaning is carried out, the drying is carried out, the sampling is carried out for GDMS detection, and 99.99990% -99.99999% of ultra-high purity indium is obtained;
s3) establishing a vacuum distillation high-purity indium production database: according to the method of the step S2, high-purity indium is subjected to vacuum distillation by setting process parameters of different vacuum distillation temperatures and vacuum distillation time, the purity of the ultra-high purity indium obtained under corresponding conditions is measured, and data are input into a structured database;
s4) utilizing a machine learning method to assist in resolving the vacuum distillation process: taking the process parameters of vacuum distillation temperature, vacuum distillation time and total impurity content of the product in the step S3 as characteristic variables, performing training modeling by using different machine learning algorithms, comparing and evaluating different machine learning models through cross verification, and selecting a model with the minimum error from a plurality of prediction models as a final prediction model;
s5) a selected machine learning model is applied to predict and optimize experimental process parameters within a certain range, and a high-purity indium vacuum distillation purification process is optimized, wherein the method is that the process parameters including vacuum distillation temperature and vacuum distillation time are input into a final prediction model established in the step S4, the model is used for outputting the purity of the product indium under the corresponding parameters, and the parameters corresponding to the highest purity indium is selected as optimal parameters.
Preferably, the machine learning algorithm is a support vector regression method, a gradient ascending regression method, a Lasso regression method or an artificial neural network method.
Preferably, the different algorithms use the average absolute error and the correlation coefficient to finally obtain the optimal machine learning model.
Compared with the prior art, the invention has the following beneficial effects:
according to the invention, the purification of the metal indium is realized through secondary vacuum distillation, parameters such as vacuum distillation temperature, vacuum distillation time, vacuum degree, vacuum distillation heat preservation time and the like of equipment can be monitored in real time, data can be visualized through a computer, the data is collected and normalized to construct a database, and experimental parameters of a high-purity indium vacuum distillation process are searched and optimized by combining data in the database with various machine learning models, so that stable production of high-purity indium products with more than 6N is realized.
According to the invention, a vacuum distillation optimization model is established by means of a machine learning method, the accuracy of the vacuum distillation model is checked through experimental data and related database data, and vacuum distillation experimental technological parameters are predicted and optimized, so that the optimization of the high-purity indium vacuum distillation purification process is realized, the purity of the high-purity indium can be improved, production technological parameters are solidified, the production controllability is enhanced, the production efficiency is improved, and high-quality high-purity metal materials are provided for the semiconductor industry.
Drawings
FIG. 1 is a schematic diagram of a machine learning optimized vacuum distillation process for producing high purity indium;
FIG. 2 is a flow chart of vacuum distillation experiments and machine learning predictions.
Detailed Description
In order to more clearly illustrate the present invention, the present invention will be further described with reference to preferred embodiments and the accompanying drawings. It is to be understood by persons skilled in the art that the following detailed description is illustrative and not restrictive, and that this invention is not limited to the details given herein.
Example 1
Referring to fig. 1, the method for preparing high-purity indium by machine learning optimization vacuum distillation relates to a method for preparing high-purity indium by using a vacuum distillation furnace 4, a Dan Yingcheng material tray 6 and a Dan Yingcheng material residual liquid phase bottom tray 8 which are arranged in the vacuum distillation furnace, an electric control device 11 for controlling the vacuum distillation furnace, a power supply module 1 with a wire connection port 2 and a computer 3, wherein a high-temperature-resistant glass window for observing the inside of the furnace is arranged on a shell outside the furnace body of the vacuum distillation furnace. The vacuum distillation furnace, the electric control device, the power supply module and the computer are all devices in the prior art.
The method for preparing high-purity indium by optimizing vacuum distillation based on machine learning comprises the following specific steps:
s1) firstly purifying indium at different distillation temperatures, and simultaneously realizing high-purity indium vacuum distillation experiments under different conditions by setting different distillation temperature parameters. 5N indium prepared by an electrolysis method is selected as a raw material: the electrolytic method of the prior art is adopted to prepare 5N indium. The electrolyte system is an indium sulfate solution containing additives of sodium chloride and gelatin, the electrolyte is an electrolyte in the prior art, under the electrolysis conditions of pH 1.5-3.0, current density 30-85A/m < 2 >, and electrolysis temperature 20-30 ℃, 3-4N metal indium is melted, cast and molded to be used as an anode, a titanium plate is used as a cathode plate for electrolysis to obtain metal indium, and the metal indium is cleaned and dried to obtain 5N indium;
s2) putting 5N indium into a vacuum distillation furnace for twice vacuum distillation purification: every time, 10 kg of 5N indium is placed in a crucible 5 in a vacuum distillation furnace, a power supply and control device 11 is started, a control cabinet button 9 is opened, parameter setting on a panel 10 is carried out, and distillation temperature and distillation time are set. The interface 2 of the power module 1 can be connected with the computer 3 for remote control and data transmission, thereby performing machine learning optimization control. Vacuumizing to reduce the vacuum degree to 10 -1 When Pa is lower, the distillation temperature is set to 800 ℃, 900 ℃, 1000 ℃, 1100 ℃, 1200 ℃ and the distillation is carried out for 3 hours, 4 hours, 5 hours, 6 hours, 7 hours and 8 hours at each set temperature. Liquid phase bottom tray for taking quartz material residue after each distillation8, performing secondary distillation at 1200-1500 ℃ for 5-8 h, closing a heating program and a vacuumizing system after the heating and heat preservation stage is finished, cooling to room temperature along with a furnace, sampling components volatilized to a liquid phase bottom plate 8 of a quartz material residue at the top in vacuum distillation, performing ultrasonic cleaning, drying, sampling, performing GDMS detection, and obtaining ultra-high purity indium with the concentration of 6N or above;
s3) establishing a vacuum distillation high-purity indium production database: according to the method of the step S2, high-purity indium is subjected to vacuum distillation by setting process parameters of different vacuum distillation temperatures and vacuum distillation time, the purity of the ultra-high purity indium obtained under corresponding conditions is measured, and data are input into a structured database;
s4) utilizing a machine learning method to assist in resolving the vacuum distillation process: taking the process parameters of vacuum distillation temperature and vacuum distillation time in the step S3 as characteristic variables, taking the total impurity content of the product as a target variable, performing training modeling by using different machine learning algorithms, comparing and evaluating different machine learning models through cross verification, and selecting a model with the minimum error from a plurality of prediction models as a final prediction model;
s5) a selected machine learning model is applied to predict and optimize experimental process parameters within a certain range, and a high-purity indium vacuum distillation purification process is optimized, wherein the method is that the process parameters including vacuum distillation temperature and vacuum distillation time are input into a final prediction model established in the step S4, the model is used for outputting the purity of the product indium under the corresponding parameters, and the parameters corresponding to the highest purity indium is selected as optimal parameters.
Example 2
The method for purifying indium under different experimental conditions comprises the following steps: preparation of 5N indium according to step S1 of example 1, 10 kg of 5N indium was placed in a graphite crucible and evacuated to 10 -3 Vacuum distillation is carried out below pa, the first distillation temperature is respectively set to 800-1200 ℃, and the first distillation time is set to 3-6 h. The residual liquid phase bottom component is taken after the first distillation and subjected to a second vacuum distillation (vacuum degree is unchanged) at 800-1000 ℃ for 5-6 h. After the heating and heat preserving stage is finished, the heating program and the pumping are closedAnd (3) performing GDMS detection by sampling after ultrasonic cleaning and drying of a vacuum system to obtain high-purity indium with the purity of 6N and above.
Example 3
The high purity indium prepared according to the above example 2 was continuously subjected to a machine learning optimized vacuum distillation process as follows:
1. establishing a vacuum distillation high-purity indium production database: based on production data at different temperatures and different distillation times, the data includes high purity indium vacuum distillation process parameters: the first distillation time, the first distillation temperature, the second distillation time, the second distillation temperature and the target value of high purity indium purity, and the production data are stored in a database with a structure shown in table 1;
table 1 operating process parameters and purity database for vacuum distillation process products
2. The machine learning method is utilized to excavate data, and the auxiliary analysis vacuum distillation process is adopted: and (3) taking the technological parameters in the step (S3) such as the vacuum distillation temperature and the vacuum distillation time as characteristic variables, taking the total impurity content of the product as a target variable, performing training modeling by using different machine learning algorithms, comparing and evaluating different machine learning models through cross verification, and selecting the model with the minimum error from multiple prediction models as a final prediction model. The specific method is to perform machine learning prediction based on the database of the table 1, and perform high-purity indium machine learning modeling and application according to the flow shown in fig. 2. 75% of the data in the database were randomly extracted as training data, the remaining 25% as test data. Four commonly used machine learning regression algorithms are adopted, including support vector regression (support vector regression, SVR), gradient ascending regression (gradient boosting regression, GBR), lasso regression and artificial neural network (Artificial Neural Network, ANN) are adopted to model training sets respectively, the optimal super parameters of the algorithm are searched by a lattice point search method, a five-fold cross validation evaluation model is adopted to calculate the average absolute error (mean average error, MAE) of the five-fold cross validation respectively, and a machine learning algorithm model with the minimum average absolute error is selected as a final prediction model. The machine learning model constructed by the GBR algorithm shown in table 2 has the smallest mean absolute error.
Table 2 different machine learning algorithm performances
| Machine learning algorithm | Mean Absolute Error (MAE) | Correlation coefficient (R) |
| SVR | 0.000015 | 0.97 |
| GBR | 0.000008 | 0.99 |
| Lasso | 0.000027 | 0.94 |
| ANN | 0.000025 | 0.95 |
3. And a machine learning model constructed by GBR algorithm is used for predicting new experimental parameter combinations, optimizing a high-purity indium vacuum distillation purification process and being used for actual production experiments. The specific method is that the vacuum distillation temperature and the vacuum distillation time are input into a machine learning model constructed by the GBR algorithm established in the step S4, different distillation time and temperature are combined, the purity of the product is predicted, the predicted result is shown in a table 3, and the parameter with the production number of 48 is the optimal parameter.
TABLE 3 model predictive results for vacuum distillation process products
The method is not only suitable for preparing high-purity indium based on machine learning optimization vacuum distillation, but also suitable for optimizing other similar metal purification processes, including but not limited to tellurium, cadmium, arsenic and the like.
The percentages stated in the present invention are mass percentages unless otherwise indicated.
Claims (3)
1. The method for preparing the high-purity indium by optimizing vacuum distillation based on machine learning is characterized by comprising the following steps of:
s1) preparing 5N indium by using an electrolysis method as a raw material: the electrolyte system of the electrolytic method is indium sulfate solution containing additives of sodium chloride and gelatin, the pH value is 1.5-3.0, and the current density is 30-85A/m 2 Under the electrolysis condition that the electrolysis temperature is 20-30 ℃, 3-4N metal indium is melted and cast to be used as an anode, a titanium plate is used as a cathode plate for electrolysis to obtain metal indium, and the metal indium is cleaned and dried to obtain 5N indium;
s2) putting 5N indium into a vacuum distillation furnace for vacuum distillation and twice purification: placing 5N indium into a crucible of a vacuum distillation furnace, vacuumizing, and reducing the vacuum degree to 10 -1 Under Pa, the temperature is set to 800-1200 ℃ for one time distillation for 3-8 h, then the residual liquid phase bottom component is taken for secondary distillation at 1200-1500 ℃ for 5-8 h, after the heating and heat preservation stage is finished, the heating program and the vacuumizing system are closed, the furnace is cooled to room temperature, the component volatilized from the top in the vacuum distillation is sampled, the ultrasonic cleaning and drying are carried out, and the sampling is carried out GDMS detection to obtain 99.99990%About 99.99999% of ultra-high purity indium;
s3) establishing a vacuum distillation high-purity indium production database: according to the method of the step S2, high-purity indium is subjected to vacuum distillation by setting process parameters of different vacuum distillation temperatures and vacuum distillation time, the purity of the ultra-high purity indium obtained under corresponding conditions is measured, and data are input into a structured database;
s4) utilizing a machine learning method to assist in resolving the vacuum distillation process: taking the process parameters of vacuum distillation temperature and vacuum distillation time in the step S3 as characteristic variables, taking the total impurity content of the product as a target variable, performing training modeling by using different machine learning algorithms, comparing and evaluating different machine learning models through cross verification, and selecting a model with the minimum error from a plurality of prediction models as a final prediction model;
s5) a selected machine learning model is applied to predict and optimize experimental process parameters within a certain range, and a high-purity indium vacuum distillation purification process is optimized, wherein the method is that the process parameters including vacuum distillation temperature and vacuum distillation time are input into a final prediction model established in the step S4, the model is used for outputting the purity of the product indium under the corresponding parameters, and the parameters corresponding to the highest purity indium is selected as optimal parameters.
2. The method for preparing high-purity indium based on machine learning optimization vacuum distillation according to claim 1, wherein the machine learning algorithm is a support vector regression method, a gradient rising regression method, a Lasso regression method or an artificial neural network method.
3. The method for preparing high-purity indium based on machine learning optimization vacuum distillation according to claim 1, wherein the different algorithms use average absolute errors and correlation coefficients to finally obtain an optimal machine learning model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210487994.9A CN114934198B (en) | 2022-05-06 | 2022-05-06 | Method for preparing high-purity indium based on machine learning optimization vacuum distillation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210487994.9A CN114934198B (en) | 2022-05-06 | 2022-05-06 | Method for preparing high-purity indium based on machine learning optimization vacuum distillation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114934198A CN114934198A (en) | 2022-08-23 |
| CN114934198B true CN114934198B (en) | 2024-03-01 |
Family
ID=82864140
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210487994.9A Active CN114934198B (en) | 2022-05-06 | 2022-05-06 | Method for preparing high-purity indium based on machine learning optimization vacuum distillation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114934198B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118170104B (en) * | 2024-04-02 | 2024-10-01 | 太和气体（荆州）有限公司 | Propylene production control system based on data analysis and optimization method thereof |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1490433A (en) * | 2002-10-16 | 2004-04-21 | 沈奕林 | Preparation of high purity indium |
| CN102031372A (en) * | 2010-12-01 | 2011-04-27 | 长春工业大学 | Online prediction system for terminal composition of low-carbon ferrochromium in silicothermic smelting |
| CN103103566A (en) * | 2013-01-31 | 2013-05-15 | 湖南化工研究院 | Preparation method of high-purity indium |
| CN109735724A (en) * | 2019-02-18 | 2019-05-10 | 云南锡业集团（控股）有限责任公司研发中心 | The method for preparing 6N high purity indium is once evaporated in vacuo |
| CN113802151A (en) * | 2021-10-27 | 2021-12-17 | 上海大学 | Multi-channel electrolysis device for high-purity indium electrorefining and machine learning process optimization method |
| CN114117884A (en) * | 2021-10-21 | 2022-03-01 | 东北大学 | Intelligent sintering ore blending system based on big data before iron |
-
2022
- 2022-05-06 CN CN202210487994.9A patent/CN114934198B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1490433A (en) * | 2002-10-16 | 2004-04-21 | 沈奕林 | Preparation of high purity indium |
| CN102031372A (en) * | 2010-12-01 | 2011-04-27 | 长春工业大学 | Online prediction system for terminal composition of low-carbon ferrochromium in silicothermic smelting |
| CN103103566A (en) * | 2013-01-31 | 2013-05-15 | 湖南化工研究院 | Preparation method of high-purity indium |
| CN109735724A (en) * | 2019-02-18 | 2019-05-10 | 云南锡业集团（控股）有限责任公司研发中心 | The method for preparing 6N high purity indium is once evaporated in vacuo |
| CN114117884A (en) * | 2021-10-21 | 2022-03-01 | 东北大学 | Intelligent sintering ore blending system based on big data before iron |
| CN113802151A (en) * | 2021-10-27 | 2021-12-17 | 上海大学 | Multi-channel electrolysis device for high-purity indium electrorefining and machine learning process optimization method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114934198A (en) | 2022-08-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114934198B (en) | Method for preparing high-purity indium based on machine learning optimization vacuum distillation | |
| Zhang et al. | Local parameter optimization of LSSVM for industrial soft sensing with big data and cloud implementation | |
| CN113802151B (en) | Multi-channel electrolysis device for high-purity indium electrorefining and machine learning process optimization method | |
| CN113094860B (en) | Industrial control network flow modeling method based on attention mechanism | |
| CN111829688A (en) | Temperature field distribution prediction method in working process of lithium ion battery | |
| CN116384700A (en) | Energy efficiency optimization method for central air conditioning system based on data mining | |
| CN112700085A (en) | Association rule based method, system and medium for optimizing steady-state operation parameters of complex system | |
| CN117668743B (en) | Time sequence data prediction method of association time-space relation | |
| Polekhina et al. | Identification of magnetic field strength realisation as a necessary solution for high-quality metal synthesis | |
| CN116629416A (en) | Photovoltaic power station power prediction method and device | |
| CN117540505A (en) | Simulation method for precision forming of lost foam | |
| Yang et al. | FARIMA Model Based on Particle Swarm-genetic Hybrid Algorithm Optimization and Application | |
| CN114774990B (en) | Energy-saving-oriented copper electrolysis process parameter optimization method | |
| CN114692341B (en) | Fan fault prediction method based on fused time-space diagram ConvLSTM model and storage medium | |
| CN112231977A (en) | Construction method of formula parameter library of redrying machine | |
| CN117577229A (en) | Optimization method and optimization system for preparing high-purity metal through vacuum distillation | |
| CN114807635B (en) | Method for preparing high-purity indium by multi-channel array directional solidification based on machine learning | |
| CN116705211B (en) | Digital twin-based online prediction method and system for copper loss rate of oxygen-enriched copper molten pool | |
| CN115583654B (en) | Polysilicon reduction furnace current control method based on imitation learning | |
| Yao et al. | Data-driven Methods for Quality Prediction of Aluminum Alloy Ingots | |
| CN117954018A (en) | Method for optimizing multi-metal monocrystal purification based on transfer learning | |
| CN117684259A (en) | Method for optimizing PVT method SiC crystal growth based on machine learning | |
| CN116837422B (en) | Production process and system of high-purity oxygen-free copper material | |
| CN118395880B (en) | Intermittent process temperature prediction method of multi-stage fusion cyclic neural network | |
| CN107991875A (en) | A kind of control method of the Chemical Batch Process enhancing model based on fail-safe analysis |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |