A method for optimizing the preparation formula of polystyrene microspheres and a preparation device

Abstract

The present invention discloses a method and device for optimizing the preparation formula of polystyrene microspheres, and relates to the technical field of polymer material preparation. The present invention provides a method for optimizing the preparation formula of polystyrene microspheres, which can realize the inverse design of the formula through Bayesian optimization and the established microsphere size neural network prediction model, effectively reducing the number of data sets collected, and optimizing the formula through iterative design, so that the formula corresponding to the corresponding microsphere size can be obtained more quickly. The present invention also provides a device for preparing polystyrene microspheres, which can automatically complete the preparation process by executing the polystyrene microsphere preparation program and the cleaning program through a controller, and can realize the preparation of microspheres of multiple batches of formulas by cooperating with existing storage tank replacement equipment or setting a multi-channel electronic valve. When collecting data sets, batch collection can be automatically performed to save manpower.

Description

Technical Field

[0001] The invention relates to the technical field of polymer material preparation, and in particular to a polystyrene microsphere preparation formula optimization method and a preparation device. Background Art

[0002] As an important polymer material, polystyrene microspheres are widely used in many fields due to their unique physical and chemical properties. Although traditional polystyrene microsphere preparation methods, such as suspension polymerization, emulsion polymerization, and dispersion polymerization, can produce microspheres that meet certain requirements, they generally have problems such as complex operation, low production efficiency, and poor product stability. Especially in terms of microsphere particle size control, morphology regulation, and dispersibility, traditional preparation methods often fail to achieve the desired results. In addition, with the advancement of science and technology and the development of industry, higher automation and intelligent requirements are put forward for the preparation of polystyrene microspheres.

[0003] Moreover, the particle size and quality of polystyrene microspheres are difficult to control. There is no linear law, and it is difficult to predict the formula data of the required particle size and quality from complex experimental data.

[0004] Microsphere synthesis is aimed at a practical problem. The existing technology uses machine learning algorithms to predict formulas, which often requires a large amount of data sets to train the model, while the collection of training sets is long and difficult. Summary of the invention

[0005] In view of the above-mentioned deficiencies in the prior art, the present invention provides a method for optimizing the preparation formula of polystyrene microspheres and a preparation device.

[0006] In order to achieve the above-mentioned object of the invention, the technical solution adopted by the present invention is:

[0007] A method for optimizing the preparation formula of polystyrene microspheres is provided, which comprises the following steps:

[0008] S1: Randomly generate 30 recipes, use the 30 recipes to prepare polystyrene microspheres, measure the size of the prepared polystyrene microspheres, integrate the recipes and the corresponding polystyrene microsphere sizes and standard deviation coefficients to obtain a data set;

[0009] S2: Establish a Bayesian optimization model and use the data set to train the Bayesian optimization model;

[0010] S3: Establish a microsphere size neural network prediction model based on the DNN model, and use the data set to train the microsphere size neural network prediction model;

[0011] The microsphere size neural network prediction model has a four-layer structure, with 4 neurons in the first layer; 100 neurons in the second and third layers, and 2 neurons in the fourth layer, which output the predicted size and the predicted standard deviation coefficient respectively;

[0012] S4: Set the target size and input the target size into the trained Bayesian optimization model. The Bayesian optimization model generates a prediction recipe through the gp_hedge strategy.

[0013] Perform iterative operations: input the prediction formula into the trained microsphere size neural network prediction model to obtain the iterative prediction size and iterative prediction standard deviation coefficient; the Bayesian optimization model generates an iterative prediction formula through the gp_hedge strategy based on the iterative prediction size and iterative prediction standard deviation coefficient;

[0014] After ten iterations, enter S5;

[0015] S5: using the predicted formula outputted from the tenth iteration to prepare polystyrene microspheres, measuring the size and standard deviation coefficient of the polystyrene microspheres prepared by the predicted formula, and outputting the formula when the difference Δl between the actual diameter of the polystyrene microspheres and the target diameter is less than 20 nm and the standard deviation coefficient cv is less than 3%;

[0016] Otherwise, the predicted formula output from the tenth iteration was reused to prepare polystyrene microspheres;

[0017] After ten preparations, if the difference Δl between the actual diameter and the target size is still not less than 20nm and the standard deviation coefficient cv is still not less than 3%, the formulas prepared ten times and the obtained polystyrene microsphere size and standard deviation coefficient are integrated, returned to S1 and added to the data set, and S2 is re-executed until the difference Δl between the actual diameter of the polystyrene microspheres and the target size is less than 20nm and the standard deviation coefficient cv is less than 3%, then the formula is output and the cycle is exited.

[0018] Furthermore, in step S3, before training the microsphere size neural network prediction model, the data set is expanded by:

[0019] Perform data perturbation on the formula values, and perform the P operation on the values ​​of potassium persulfate, sodium dodecyl sulfate, pure water, and ethanol in any formula in the data set, where P is randomly selected from (-1, 0, 1);

[0020] The expanded 120 data were used to train the microsphere size neural network prediction model.

[0021] Furthermore, the adam optimizer was used to train the microsphere size neural network prediction model, and the period epoch was set to 120, and the batch was set to 1; and the activation function of the microsphere size neural network prediction model was relu, and the loss function was the mean square error loss function MSE.

[0022] Furthermore, in step S2, the Optimizer optimizer is used to train the Bayesian optimization model, and gp_hedge and Gaussian process GP are set as hyperparameters;

[0023] The optimization conditions are: diameter difference Δl = |size-object| < 20 nm and standard deviation coefficient cv < 3%;

[0024] The optimization function is:

[0025] min{value}=(size-object|+bias)*transform(cv);

[0026]

[0027] Among them, value is the optimized value, that is, the formula parameter; object is the target size, size is the predicted size of the microsphere size neural network prediction model or the actual size of the prepared polystyrene microspheres; bias is a constant, and bias=1.

[0028] Furthermore, the range of the randomly generated formula and the search space range of the Bayesian optimization model for the preparation formula of polystyrene microspheres are: potassium persulfate [10,40], sodium dodecyl sulfate [10,40], pure water [6,23], ethanol [5,40], styrene [8,12], and the value of any raw material is an integer.

[0029] Also provided is a preparation device applied to the above-mentioned polystyrene microsphere preparation formula optimization method, comprising a reaction bottle, the reaction bottle is provided with a nitrogen input end, a raw material input end and a finished product output end, the nitrogen input end is provided with an electric valve, and the nitrogen input end is connected to a nitrogen storage bottle;

[0030] An electronic multi-channel valve is provided at the raw material input end, and the electronic multi-channel valve is respectively connected to a number of raw material storage tanks, and an input peristaltic pump is provided between any raw material storage tank and the electronic multi-channel valve;

[0031] The finished product output end is connected to a finished product storage tank, an output peristaltic pump is arranged between the finished product storage tank and the finished product output end, and an electric valve is also arranged at the finished product output end;

[0032] A pH detector is also provided in the reaction bottle;

[0033] A water bath heating device is also provided, and the reaction bottle is located in the water bath heating zone of the water bath heating device;

[0034] Also included is a mobile terminal, wherein the mobile terminal has a built-in controller, and the controller is electrically connected to the pH detector, the electric valve, the electronic multi-channel valve, a plurality of input peristaltic pumps, and an output peristaltic pump;

[0035] The mobile terminal has a built-in polystyrene microsphere preparation program; the polystyrene microsphere preparation program prepares polystyrene microspheres according to a formula obtained by a polystyrene microsphere preparation formula optimization method.

[0036] Furthermore, it also includes a fume hood, and the reaction bottle, finished product storage tank and waste liquid storage tank are all located in the fume hood.

[0037] Further, the cleaning component includes an ultrasonic oscillator, which is used to oscillate the liquid in the reaction bottle;

[0038] The reaction bottle is also provided with a waste liquid output end, which is connected to a waste liquid storage tank, a waste liquid peristaltic pump is provided between the waste liquid storage tank and the waste liquid output end, and an electric valve is also provided at the waste liquid output end; the ultrasonic oscillator, the waste liquid peristaltic pump and the electric valve are all electrically connected to the controller.

[0039] Furthermore, the steps of the polystyrene microsphere preparation procedure are as follows:

[0040] P1: The electric valve at the nitrogen input end is opened to turn the reaction bottle into a nitrogen environment, and nitrogen is continuously introduced into the reaction bottle;

[0041] P2: The water bath heating device heats the temperature in the reaction bottle to 40°C;

[0042] P3: Control any channel of the electronic multi-channel valve to open in sequence, and introduce potassium persulfate solution, sodium dodecyl sulfate solution, ultrapure water, ethanol and styrene into the reaction bottle in sequence through the input peristaltic pump according to the formula obtained by the polystyrene microsphere preparation formula optimization method;

[0043] P4: The water bath heating device heats the temperature in the reaction bottle to 70°C; the pH detector continuously detects the pH value of the liquid in the reaction bottle. When the pH value stabilizes, the reaction is completed, the water bath heating device stops heating, the electric valve at the finished product output end opens, and the output peristaltic pump pumps the finished product in the reaction bottle to the finished product storage tank; the electric valve at the finished product output end closes, completing the polystyrene microsphere preparation procedure.

[0044] Furthermore, the controller also has a built-in cleaning program, and the specific steps of the cleaning program are as follows:

[0045] B1: After the input peristaltic pump introduces ultrapure water into the reaction bottle, the ultrasonic oscillator oscillates and cleans the reaction bottle; after the oscillation is completed, the electric valve at the waste liquid output end opens, the waste liquid peristaltic pump pumps the contaminated ultrapure water to the waste liquid storage tank, and then the electric valve at the waste liquid output end closes;

[0046] B2: After the input peristaltic pump introduces ethanol into the reaction bottle, the ultrasonic oscillator oscillates and cleans the reaction bottle; after the oscillation is completed, the electric valve at the waste liquid output end opens, the waste liquid peristaltic pump pumps the contaminated ethanol to the waste liquid storage tank, and then the electric valve at the waste liquid output end closes;

[0047] B3: After the input peristaltic pump introduces ultrapure water into the reaction bottle again, the ultrasonic oscillator oscillates and cleans the reaction bottle; after the oscillation is completed, the electric valve at the waste liquid output end opens, and the waste liquid peristaltic pump pumps the contaminated ultrapure water to the waste liquid storage tank, and then the electric valve at the waste liquid output end is closed to complete the cleaning procedure.

[0048] The beneficial effects of the present invention are:

[0049] The present invention provides a method for optimizing the preparation formula of polystyrene microspheres. The inverse design of the formula can be achieved through Bayesian optimization and an established neural network prediction model for microsphere size, which effectively reduces the number of data sets collected. The formula is optimized through iterative design, and the formula corresponding to the corresponding microsphere size can be obtained more quickly.

[0050] The present invention also provides a polystyrene microsphere preparation device. The preparation device can automatically complete the preparation process by executing the polystyrene microsphere preparation program and the cleaning program through a controller. In conjunction with existing storage tank replacement equipment or the setting of a multi-channel electronic valve, the microsphere preparation of multiple batches of formulas can be achieved. When collecting data sets, batch collection can be automatically performed to save manpower. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 Schematic diagram of the process for optimizing the formulation of polystyrene microspheres;

[0052] Figure 2 It is a schematic diagram of the structure of the device for preparing polystyrene microspheres;

[0053] Among them, 1. reaction bottle; 2. nitrogen storage bottle; 3. electronic multi-channel valve; 4. raw material storage tank group; 5. input peristaltic pump; 6. finished product storage tank; 7. output peristaltic pump; 8. water bath heating device; 9. mobile terminal; 10. fume hood; 11. ultrasonic oscillator; 12. waste liquid storage tank; 13. waste liquid peristaltic pump. DETAILED DESCRIPTION

[0054] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0055] Example 1

[0056] like Figure 1 As shown, a method for optimizing the preparation formula of polystyrene microspheres comprises the following steps:

[0057] S1: Randomly generate 30 recipes, use each of the 30 recipes to prepare polystyrene microspheres, measure the size of the prepared polystyrene microspheres, integrate the recipes and the corresponding size of the generated polystyrene microspheres and the standard deviation coefficient to obtain a data set; in the specific implementation, use Python's random module random to generate 30 recipes.

[0058] S2: Establish a Bayesian optimization model and use the data set to train the Bayesian optimization model;

[0059] S3: Establish a microsphere size neural network prediction model based on the DNN model, and use the data set to train the microsphere size neural network prediction model;

[0060] The microsphere size neural network prediction model has a four-layer structure, with 4 neurons in the first layer; 100 neurons in the second and third layers, and 2 neurons in the fourth layer, which output the predicted size and the predicted standard deviation coefficient respectively;

[0061] S4: Set the target size and input the target size into the trained Bayesian optimization model. The Bayesian optimization model generates a prediction recipe through the gp_hedge strategy.

[0062] Perform iterative operations: input the prediction formula into the trained microsphere size neural network prediction model to obtain the iterative prediction size and iterative prediction standard deviation coefficient; the Bayesian optimization model generates an iterative prediction formula through the gp_hedge strategy based on the iterative prediction size and iterative prediction standard deviation coefficient;

[0063] After ten iterations, enter S5;

[0064] S5: using the predicted formula outputted from the tenth iteration to prepare polystyrene microspheres, measuring the size and standard deviation coefficient of the polystyrene microspheres prepared by the predicted formula, and outputting the formula when the difference Δl between the actual diameter of the polystyrene microspheres and the target diameter is less than 20 nm and the standard deviation coefficient cv is less than 3%;

[0065] Otherwise, the predicted formula output from the tenth iteration was reused to prepare polystyrene microspheres;

[0066] After ten preparations, if the difference Δl between the actual diameter and the target size is still not less than 20nm and the standard deviation coefficient cv is still not less than 3%, the formulas prepared ten times and the obtained polystyrene microsphere size and standard deviation coefficient are integrated, returned to S1 and added to the data set, and S2 is re-executed until the difference Δl between the actual diameter of the polystyrene microspheres and the target size is less than 20nm and the standard deviation coefficient cv is less than 3%, then the formula is output and the cycle is exited.

[0067] Step S3: Before training the microsphere size neural network prediction model, the data set is expanded by:

[0068] Perform data perturbation on the formula values, and perform the P operation on the values ​​of potassium persulfate, sodium dodecyl sulfate, pure water, and ethanol in any formula in the data set, where P is randomly selected from (-1, 0, 1);

[0069] The expanded 120 data were used to train the microsphere size neural network prediction model.

[0070] The adam optimizer was used to train the microsphere size neural network prediction model, and the period epoch was set to 120 and the batch was set to 1. The activation function of the microsphere size neural network prediction model was relu, and the loss function was the mean square error loss function MSE.

[0071] Step S4 uses the Optimizer optimizer to train the Bayesian optimization model, and sets gp_hedge and Gaussian process GP as hyperparameters;

[0072] The optimization conditions are: diameter difference Δl = |size-object| < 20 nm and standard deviation coefficient cv < 3%;

[0073] The optimization function is:

[0074] min{value}=(size-object|+bias)*transform(cv);

[0075]

[0076] Among them, value is the optimized value, that is, the formula parameter; object is the target size, size is the predicted size of the microsphere size neural network prediction model or the actual size of the prepared polystyrene microspheres; bias is a constant, and bias=1.

[0077] The range of randomly generated formulas and the search space range of the Bayesian optimization model for the preparation formula of polystyrene microspheres are: potassium persulfate [10,40], sodium dodecyl sulfate [10,40], pure water [6,23], ethanol [5,40], styrene [8,12], and the value of any raw material is an integer; in specific implementation, in any formula, potassium persulfate and sodium dodecyl sulfate are solutions, and the concentration is 0.005 g / mL.

[0078] Example 2

[0079] like Figure 2 As shown, a preparation device applied to the above-mentioned polystyrene microsphere preparation formula optimization method includes a reaction bottle 1, the reaction bottle 1 is provided with a nitrogen input end, a raw material input end and a finished product output end, the nitrogen input end is provided with an electric valve, and the nitrogen input end is connected to a nitrogen storage bottle 2 through a pipeline;

[0080] An electronic multi-channel valve 3 is provided at the raw material input end, and the electronic multi-channel valve 3 is connected to a plurality of raw material storage tanks through pipelines, and an input peristaltic pump 5 is provided between any raw material storage tank and the electronic multi-channel valve 3;

[0081] The finished product output end is connected to a finished product storage tank 6 through a pipeline, an output peristaltic pump 7 is provided between the finished product storage tank 6 and the finished product output end, and an electric valve is also provided at the finished product output end;

[0082] A pH detector is also provided in the reaction bottle 1;

[0083] A water bath heating device 8 is also provided, and the reaction bottle 1 is located in the water bath heating zone of the water bath heating device 8;

[0084] It also includes a mobile terminal 9, which has a built-in controller, and the controller is electrically connected to the pH detector, the electric valve, the electronic multi-channel valve 3, a plurality of input peristaltic pumps 5 and an output peristaltic pump 7;

[0085] The mobile terminal 9 has a built-in polystyrene microsphere preparation program; the polystyrene microsphere preparation program prepares polystyrene microspheres according to a formula obtained by a polystyrene microsphere preparation formula optimization method.

[0086] It also includes a fume hood 10, in which the reaction bottle 1, the finished product storage tank 6 and the waste liquid storage tank 12 are all located. In the specific implementation, there are five raw material storage tanks, which respectively store potassium persulfate solution, sodium dodecyl sulfate solution, ultrapure water, ethanol and styrene. The five raw material storage tanks constitute a raw material storage tank group 4 and are placed in a cabinet below the fume hood 10.

[0087] The cleaning component includes an ultrasonic oscillator 11, which is used to oscillate the liquid in the reaction bottle 1;

[0088] The reaction bottle 1 is also provided with a waste liquid output end, which is connected to a waste liquid storage tank 12 through a pipeline, a waste liquid peristaltic pump 13 is provided between the waste liquid storage tank 12 and the waste liquid output end, and an electric valve is also provided at the waste liquid output end; the ultrasonic oscillator 11, the waste liquid peristaltic pump 13 and the electric valve are all electrically connected to the controller.

[0089] The specific steps of the polystyrene microsphere preparation procedure are as follows:

[0090] P1: The electric valve at the nitrogen input end is opened, so that the reaction bottle 1 becomes a nitrogen environment, and nitrogen is continuously introduced into the reaction bottle 1;

[0091] P2: The water bath heating device 8 heats the temperature in the reaction bottle 1 to 40°C;

[0092] P3: Control any channel of the electronic multi-channel valve 3 to open in sequence, and introduce potassium persulfate solution, sodium dodecyl sulfate solution, ultrapure water, ethanol and styrene into the reaction bottle 1 in sequence through the input peristaltic pump 5 according to the formula obtained by the polystyrene microsphere preparation formula optimization method;

[0093] P4: The water bath heating device 8 heats the temperature in the reaction bottle 1 to 70°C; the pH detector continuously detects the pH value of the liquid in the reaction bottle 1. When the pH value is stable, the reaction is completed, the water bath heating device 8 stops heating, the electric valve at the finished product output end is opened, and the output peristaltic pump 7 pumps the finished product in the reaction bottle 1 to the finished product storage tank 6; the electric valve at the finished product output end is closed, and the polystyrene microsphere preparation procedure is completed. When the polystyrene microsphere preparation procedure is executed, the fume hood is fully opened.

[0094] The controller also has a built-in cleaning program. The specific steps of the cleaning program are as follows:

[0095] B1: After the input peristaltic pump 5 introduces ultrapure water into the reaction bottle 1, the ultrasonic oscillator 11 oscillates and cleans the reaction bottle 1; after the oscillation is completed, the electric valve at the waste liquid output end is opened, and the waste liquid peristaltic pump 13 pumps the contaminated ultrapure water to the waste liquid storage tank 12, and then the electric valve at the waste liquid output end is closed;

[0096] B2: After the input peristaltic pump 5 introduces ethanol into the reaction bottle 1, the ultrasonic oscillator 11 oscillates and cleans the reaction bottle 1; after the oscillation is completed, the electric valve at the waste liquid output end is opened, and the waste liquid peristaltic pump 13 pumps the contaminated ethanol to the waste liquid storage tank 12, and then the electric valve at the waste liquid output end is closed;

[0097] B3: After the peristaltic pump 5 is input to introduce ultrapure water into the reaction bottle 1 again, the ultrasonic oscillator 11 oscillates and cleans the reaction bottle 1; after the oscillation is completed, the electric valve at the waste liquid output end is opened, and the waste liquid peristaltic pump 13 pumps the contaminated ultrapure water to the waste liquid storage tank 12, and then the electric valve at the waste liquid output end is closed to complete the cleaning procedure.

Claims

1. A method for optimizing the preparation formula of polystyrene microspheres, characterized in that: The steps include: S1: Randomly generate 30 recipes, use the 30 recipes to prepare polystyrene microspheres, measure the size of the prepared polystyrene microspheres, integrate the recipes and the corresponding polystyrene microsphere sizes and standard deviation coefficients to obtain a data set; S2: Establish a Bayesian optimization model and use the data set to train the Bayesian optimization model; S3: Establish a microsphere size neural network prediction model based on the DNN model, and use the data set to train the microsphere size neural network prediction model; The microsphere size neural network prediction model has a four-layer structure, wherein the first layer is provided with 4 neurons; the second and third layers are provided with 100 neurons each, and the fourth layer is provided with 2 neurons, and the 2 neurons respectively output the predicted size and the predicted standard deviation coefficient; S4: Set the target size and input the target size into the trained Bayesian optimization model. The Bayesian optimization model generates a prediction recipe through the gp_hedge strategy. Perform iterative operations: input the prediction formula into the trained microsphere size neural network prediction model to obtain the iterative prediction size and iterative prediction standard deviation coefficient; the Bayesian optimization model generates an iterative prediction formula through the gp_hedge strategy based on the iterative prediction size and iterative prediction standard deviation coefficient; After ten iterations, enter S5; S5: Prepare polystyrene microspheres using the predicted formula output from the tenth iteration, measure the size and standard deviation coefficient of the polystyrene microspheres prepared by the predicted formula, and when the actual diameter of the polystyrene microspheres differs from the target diameter by ∆ l Less than 20nm and standard deviation coefficient cv When it is less than 3%, the formula is output; Otherwise, the predicted formula output from the tenth iteration was reused to prepare polystyrene microspheres; After ten preparations, the difference between the actual diameter and the target diameter is still not satisfied. l Less than 20nm and standard deviation coefficient cv When the diameter of the polystyrene microspheres is less than 3%, the formulas prepared ten times and the obtained polystyrene microsphere sizes and standard deviation coefficients are integrated, returned to S1 and added to the data set, and S2 is re-executed until the actual diameter of the polystyrene microspheres is less than the target diameter. l Less than 20nm and standard deviation coefficient cv When it is less than 3%, the formula is output and the cycle is exited.

2. The method for optimizing the preparation formula of polystyrene microspheres according to claim 1, characterized in that: In step S3, before training the microsphere size neural network prediction model, the data set is expanded by: Perform data perturbation on the formula values, and perform the P operation on the values ​​of potassium persulfate, sodium dodecyl sulfate, pure water, and ethanol in any formula in the data set, where P is randomly selected from (-1, 0, 1); The expanded 120 data were used to train the microsphere size neural network prediction model.

3. The method for optimizing the preparation formula of polystyrene microspheres according to claim 2, characterized in that: The adam optimizer is used to train the microsphere size neural network prediction model, and the period epoch=120 and the batch batch=1 are set; and the activation function of the microsphere size neural network prediction model is relu, and the loss function is the mean square error loss function MSE.

4. The method for optimizing the preparation formula of polystyrene microspheres according to claim 1, characterized in that: In step S2, the Optimizer optimizer is used to train the Bayesian optimization model, and gp_hedge and Gaussian process GP are set as hyperparameters; The optimization condition is: diameter difference and standard deviation coefficient ; The optimization function is: ； ； in, value is the optimized value, i.e., the recipe parameter; object is the target size, size is the predicted size of the microsphere size neural network prediction model or the actual size of the prepared polystyrene microspheres; bias is a constant, and bias =1.

5. The method for optimizing the preparation formula of polystyrene microspheres according to claim 1, characterized in that: The generation range of the randomly generated formula and the search space range of the Bayesian optimization model for the preparation formula of polystyrene microspheres are: potassium persulfate [10,40], sodium dodecyl sulfate [10,40], pure water [6,23], ethanol [5,40], styrene [8,12], and the value of any raw material is an integer.

6. A preparation device applied to the method for optimizing the preparation formula of polystyrene microspheres according to any one of claims 1 to 5, characterized in that: It comprises a reaction bottle, wherein the reaction bottle is provided with a nitrogen input end, a raw material input end and a finished product output end, the nitrogen input end is provided with an electric valve, and the nitrogen input end is connected to a nitrogen storage bottle; The raw material input end is provided with an electronic multi-channel valve, the electronic multi-channel valve is respectively connected to a plurality of raw material storage tanks, and an input peristaltic pump is provided between any raw material storage tank and the electronic multi-channel valve; The finished product output end is connected to a finished product storage tank, an output peristaltic pump is arranged between the finished product storage tank and the finished product output end, and an electric valve is also arranged at the finished product output end; A pH detector is also provided in the reaction bottle; A water bath heating device is also provided, and the reaction bottle is located in a water bath heating zone of the water bath heating device; Also included is a mobile terminal, wherein the mobile terminal has a built-in controller, and the controller is electrically connected to the pH detector, the electric valve, the electronic multi-channel valve, a plurality of input peristaltic pumps and an output peristaltic pump respectively; The mobile terminal has a built-in polystyrene microsphere preparation program; the polystyrene microsphere preparation program prepares polystyrene microspheres according to a formula obtained by a polystyrene microsphere preparation formula optimization method.

7. The preparation device according to claim 6, characterized in that: It also includes a fume hood, in which the reaction bottle, finished product storage tank and waste liquid storage tank are all located.

8. The preparation device according to claim 6, characterized in that: It also includes a cleaning component, which includes an ultrasonic oscillator, and the ultrasonic oscillator is used to oscillate the liquid in the reaction bottle; The reaction bottle is also provided with a waste liquid output end, the waste liquid output end is connected to a waste liquid storage tank, a waste liquid peristaltic pump is provided between the waste liquid storage tank and the waste liquid output end, and an electric valve is also provided at the waste liquid output end; the ultrasonic oscillator, the waste liquid peristaltic pump and the electric valve are all electrically connected to the controller.

9. The preparation device according to claim 6, characterized in that: The steps of the polystyrene microsphere preparation procedure are as follows: P1: The electric valve at the nitrogen input end is opened to turn the reaction bottle into a nitrogen environment, and nitrogen is continuously introduced into the reaction bottle; P2: The water bath heating device heats the temperature in the reaction bottle to 40°C; P3: Control any channel of the electronic multi-channel valve to open in sequence, and introduce potassium persulfate solution, sodium dodecyl sulfate solution, ultrapure water, ethanol and styrene into the reaction bottle in sequence through the input peristaltic pump according to the formula obtained by the polystyrene microsphere preparation formula optimization method; P4: The water bath heating device heats the temperature in the reaction bottle to 70°C; the pH detector continuously detects the pH value of the liquid in the reaction bottle. When the pH value is stable, the reaction is completed, the water bath heating device stops heating, the electric valve at the finished product output end is opened, and the output peristaltic pump pumps the finished product in the reaction bottle to the finished product storage tank; the electric valve at the finished product output end is closed, and the polystyrene microsphere preparation procedure is completed.

10. The preparation device according to claim 8, characterized in that: The controller also has a built-in cleaning program, and the specific steps of executing the cleaning program are: B1: After the input peristaltic pump introduces ultrapure water into the reaction bottle, the ultrasonic oscillator oscillates and cleans the reaction bottle; after the oscillation is completed, the electric valve at the waste liquid output end opens, the waste liquid peristaltic pump pumps the contaminated ultrapure water to the waste liquid storage tank, and then the electric valve at the waste liquid output end closes; B2: After the input peristaltic pump introduces ethanol into the reaction bottle, the ultrasonic oscillator oscillates and cleans the reaction bottle; after the oscillation is completed, the electric valve at the waste liquid output end opens, the waste liquid peristaltic pump pumps the contaminated ethanol to the waste liquid storage tank, and then the electric valve at the waste liquid output end closes; B3: After the input peristaltic pump introduces ultrapure water into the reaction bottle again, the ultrasonic oscillator oscillates and cleans the reaction bottle; after the oscillation is completed, the electric valve at the waste liquid output end opens, and the waste liquid peristaltic pump pumps the contaminated ultrapure water to the waste liquid storage tank, and then the electric valve at the waste liquid output end is closed to complete the cleaning procedure.

Images