Sweet potato vermicelli pulping process optimization method and system based on dynamic viscosity regulation and control
By constructing a digital twin model and viscosity evaluation model for the sweet potato vermicelli pulping process, the problem of unstable viscosity control in the traditional sweet potato vermicelli pulping process was solved, achieving stability in the sweet potato vermicelli pulping process and improving product quality, and generating a control scheme with the lowest cost.
Patent Information
- Application Number
- CN202511615650.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-06
- Publication Date
- 2026-01-30
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Traditional sweet potato vermicelli pulping process lacks dynamic control over viscosity, leading to unstable production and inconsistent product quality.
The system optimizes the sweet potato vermicelli pulping process based on viscosity dynamic control. It includes a process division module, a model building module, a viscosity evaluation module, and a viscosity control module. A digital twin model is constructed and a convolutional neural network is used for viscosity evaluation and control to achieve real-time dynamic control of pulp viscosity.
This technology has improved the stability of the sweet potato vermicelli pulping process and enhanced product quality. By combining digital twin models and viscosity assessment models, it can simulate the changes in pulp viscosity under different conditions in simulation software, generate the lowest-cost control scheme, and achieve real-time and lowest-cost dynamic control.
Smart Images
Figure CN121433403A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The application relates to the technical field of process optimization, in particular to a sweet potato vermicelli beating process optimization method and system based on dynamic viscosity regulation. BACKGROUND
[0002] The beating process is a key link in the production of sweet potato vermicelli, and its effect directly affects the quality of the vermicelli, and the slurry viscosity in the beating process is an important index, which is affected by factors such as material characteristics, beating time, temperature, water addition amount, etc. Traditional beating process often lacks dynamic regulation of viscosity, resulting in unstable production process and uneven product quality. In the prior art, a technical solution for real-time monitoring of slurry viscosity and other parameters by using Internet of Things technology has been introduced to solve the drawbacks of traditional manual evaluation, but it lacks effective use of various monitoring parameters, and its regulation method is only to add water or raw materials. This method is neither controllable nor cost-effective. Therefore, the application provides a sweet potato vermicelli beating process optimization method and system based on dynamic viscosity regulation. SUMMARY
[0003] The application aims to provide a sweet potato vermicelli beating process optimization method and system based on dynamic viscosity regulation.
[0004] The application can be achieved by the following technical solution: a sweet potato vermicelli beating process optimization system based on dynamic viscosity regulation, comprising the following modules: A process division module is used to obtain the whole process of beating and divide it into different process sub-processes, and obtain the basic information and equipment data of each process sub-process corresponding processing equipment; A model construction module is used to construct a three-dimensional physical model of each processing equipment according to the basic information, and construct a beating digital twin model combined with the corresponding equipment data; A viscosity evaluation module is used to obtain adjustable parameters contained in each process sub-process, obtain slurry viscosity under dynamic changes of different adjustable parameters in the digital twin model, and construct a viscosity evaluation model according to different adjustable parameters and their corresponding slurry viscosity; A viscosity regulation module is used to obtain the predicted viscosity by using the viscosity evaluation model, judge whether viscosity regulation is needed combined with the preset target viscosity, and generate a regulation scheme by using the viscosity evaluation model.
[0005] Further, the process of obtaining the whole process of beating and dividing it into different process sub-processes comprises: The whole process refers to the whole process from raw materials to slurry, including the following process steps: raw material preparation and processing, primary mixing and sizing, high temperature gelatinization and curing, cooling and viscosity control. Different process steps correspond to different processing equipment, and the processing process of the processing equipment corresponding to different process steps is regarded as a process sub-flow.
[0006] Further, the process of acquiring the basic information and equipment data of the processing equipment corresponding to each process sub-flow respectively includes: The basic information refers to the specification parameters of the processing equipment corresponding to each process sub-flow, and the equipment data includes one type of equipment data and two types of equipment data. A one-type acquisition unit is set for each processing equipment, and a two-type acquisition unit is set for the integrated beater. The one-type equipment data of the processing equipment is collected in real time through the one-type acquisition unit, including pressure, temperature, load rate, current, voltage, and vibration amplitude. The two-type equipment data of the integrated beater is collected in real time through the two-type acquisition unit, which refers to the temperature, content, and viscosity of the slurry inside the integrated beater.
[0007] Further, the process of constructing a three-dimensional physical model according to the basic information of each processing equipment and constructing a digital twin model of beater based on the corresponding equipment data includes: A three-dimensional physical model of the corresponding process sub-flow is constructed according to the basic information of each processing equipment using a three-dimensional modeling tool. A three-dimensional simulation model is obtained by simulating the processing process of the integrated beater in different process sub-flows based on the three-dimensional physical model of the integrated beater using simulation software. The one-type equipment data of the integrated beater in different process sub-flows is uploaded to the three-dimensional simulation model for synchronization. The values of the simulation parameters in the simulation process are adjusted to obtain the temperature, content, and viscosity of the slurry in the three-dimensional simulation model under different values of the simulation parameters. When the obtained temperature, content, and viscosity are the same as the two-type equipment data of the corresponding process sub-flow at the corresponding time, the three-dimensional simulation model at this time is regarded as the digital twin model of the beater.
[0008] Further, the process of acquiring the adjustable parameters contained in each process sub-flow and obtaining the slurry viscosity under the dynamic changes of different adjustable parameters in the digital twin model includes: The adjustable parameters include starch feeding amount, water feeding amount, auxiliary material type and amount, stirring speed, heating rate, gelatinization temperature and holding time, and cooling rate. In the digital twin model, the single adjustable parameter is taken as the control object, and the other adjustable parameters are taken as non-control objects and remain unchanged. The final pulp consistency of the pulp in the integrated beater is obtained when the control object is at different values. The pulp consistency of the control object at different values and the values of the non-control objects are included in the control object parameter set of the control object at different values. The control object parameter set corresponding to the control object at different values is obtained when the other single adjustable parameter is taken as the control object and the other adjustable parameters except for it are taken as non-control objects and remain unchanged.
[0009] Further, the process of constructing the viscosity evaluation model according to different adjustable parameters and their corresponding pulp consistencies includes: According to the values of the corresponding control object and the non-control objects in the different control object parameter sets, and the final pulp consistency, a viscosity evaluation set is generated, which is divided into a training set and a test set. A convolutional neural network is constructed, and the values of the different control objects and the non-control objects in the training set are taken as the input data of the convolutional neural network, and the corresponding pulp consistencies in the training set are taken as the output data of the convolutional neural network. The training set is used to train the convolutional neural network to obtain an initial convolutional neural network, and the test set is used to verify the model of the initial convolutional neural network. The initial convolutional neural network with an output less than or equal to a preset test error threshold is output as the viscosity evaluation model.
[0010] Further, the process of obtaining a predicted consistency using the viscosity evaluation model and determining whether it needs viscosity control in combination with a preset target consistency includes: In actual application scenarios, the real-time values of the adjustable parameters in the integrated beater are obtained and input into the viscosity evaluation model to obtain the corresponding predicted consistency. The predicted consistency is compared with the preset target consistency. When the predicted consistency is not within the target consistency range, it is determined that the current process sub-flow needs viscosity control. Any adjustable parameter is taken as the control object, and the other adjustable parameters are taken as non-control objects. The viscosity evaluation model is used to obtain the predicted consistency corresponding to the control object at different values when the non-control objects remain unchanged. When the predicted consistency is within the target consistency range, the value of the control object at this time is taken as the control parameter value. Respectively acquiring the controllable parameters as the control object, the control parameter value capable of making the predicted viscosity in the target viscosity range is acquired, and the total cost required for adjusting the different control objects from the current value to the corresponding control parameter value is acquired, the control object and its control parameter value with the lowest total cost are taken as its control scheme, and it is fed back to the relevant personnel.
[0011] The sweet potato vermicelli beating process optimization method based on viscosity dynamic regulation comprises the following steps: Step S1: acquiring the whole process of beating and dividing it into different process sub-processes, and acquiring the basic information and equipment data of the corresponding processing equipment of each process sub-process; Step S2: constructing a three-dimensional physical model according to the basic information of each processing equipment, and constructing a digital twin model of beating in combination with the corresponding equipment data; Step S3: acquiring the controllable parameters contained in each process sub-process, acquiring the slurry viscosity under the dynamic change of different controllable parameters in the digital twin model, and constructing a viscosity evaluation model according to different controllable parameters and the corresponding slurry viscosity; Step S4: acquiring the predicted viscosity by using the viscosity evaluation model, judging whether viscosity regulation is needed in combination with the preset target viscosity, and generating a control scheme by using the viscosity evaluation model.
[0012] Compared with the prior art, the beneficial effects of the present application are: By dividing the beating process of sweet potato vermicelli into different sub-processes and acquiring the basic information and equipment data of the corresponding processing equipment of each sub-process, the digital twin model of the whole beating process is constructed, the beating process under different conditions can be simulated in the simulation software, the change of slurry viscosity under the condition of keeping other controllable parameters unchanged and changing only a single parameter is acquired in the digital twin model, the influence of the isolated change of different parameters on the final slurry viscosity is acquired, the viscosity evaluation model is constructed to acquire the subsequent predicted viscosity, after judging whether viscosity regulation is needed, the control scheme capable of achieving the target viscosity and having the lowest cost is acquired by using the viscosity evaluation model, and the real-time and cost-effective dynamic regulation mechanism can be realized according to the current slurry viscosity. BRIEF DESCRIPTION OF DRAWINGS
[0013] Figure 1 The module schematic diagram of the present application is shown. DETAILED DESCRIPTION
[0014] As Figure 1 shown, the sweet potato vermicelli beating process optimization system based on viscosity dynamic regulation comprises the following modules: The process division module is configured to obtain a whole process of beating and divide the whole process into different process sub-processes, and obtain basic information and equipment data of processing equipment corresponding to each process sub-process respectively; The model construction module is configured to construct a three-dimensional physical model of each processing equipment according to the basic information of the processing equipment, and construct a digital twin model of beating in combination with the corresponding equipment data; The viscosity evaluation module is configured to obtain adjustable parameters contained in each process sub-process, obtain pulp viscosity under dynamic changes of different adjustable parameters in the digital twin model respectively, and construct a viscosity evaluation model according to different adjustable parameters and corresponding pulp viscosities. The viscosity control module is configured to obtain a predicted viscosity by using the viscosity evaluation model, judge whether viscosity control is needed in combination with a preset target viscosity, and generate a control scheme by using the viscosity evaluation model.
[0015] It should be further explained that, in the specific implementation process, the process of obtaining the whole process of beating and dividing the whole process into different process sub-processes includes: The "beating" process of sweet potato vermicelli is a key link connecting starch preparation and powder forming, and its essence is to mix, paste, and mature water, starch, and auxiliary materials into uniform slurry with specific rheological properties suitable for powder forming through mechanical stirring and heat action; The whole process refers to the whole process of processing water, starch, and auxiliary materials from raw materials to slurry, which will go through different process links, including raw material preparation and processing, primary mixing and beating, high-temperature pasting and maturation, cooling and viscosity control. Different process links correspond to different processing equipment, and the processing process of the processing equipment corresponding to different process links is taken as a process sub-process respectively. Specifically, the raw material preparation and processing are used to ensure that the starch raw material input into the beating process is stable in quality and accurate in measurement, laying a foundation for the stability of the subsequent process. The processing equipment is a screw conveyor, an electronic weighing hopper, and an auxiliary material dissolving and stirring tank. The primary mixing and beating are used to mix water, starch, and auxiliary materials into uniform emulsion suspension to avoid clumping and prepare for pasting. The processing equipment is an integrated beater. The high-temperature pasting and maturation are used to heat and continuously stir to make starch particles swell and rupture by absorbing water, and to paste into thick and transparent gelatinous mature slurry. The processing equipment is an integrated beater. The cooling and viscosity control are used to cool the high-temperature slurry after maturation to a process temperature suitable for subsequent powder forming operation, and to accurately adjust the viscosity to a target range during the process. The processing equipment is an integrated beater and a discharge pump.
[0016] It should be further explained that, in the specific implementation process, the process of obtaining the basic information and equipment data of the processing equipment corresponding to each process sub-flow includes: The basic information refers to the specifications and parameters of the processing equipment corresponding to each process sub-process, including all relevant data necessary for building a three-dimensional physical model. In actual application scenarios, a first-class acquisition unit is set up for each processing equipment, and a second-class acquisition unit is set up for the integrated pulping machine. The first type of acquisition unit collects first-type equipment data of the processing equipment in real time, and the second type of acquisition unit collects second-type equipment data of the integrated pulping machine in real time. The equipment data includes first-type equipment data and second-type equipment data. The first type of equipment data refers to the pressure, temperature, load rate, current, voltage, and vibration amplitude of each processing device. The first type of equipment data is not completely the same for different processing devices. The second type of equipment data refers to the temperature, content, and viscosity of the pulp inside the integrated pulping machine at the end of each process sub-process.
[0017] It should be further explained that, in the specific implementation process, the process of constructing a three-dimensional physical model of each processing equipment based on its basic information, and then constructing a digital twin model of the pulping process by combining the corresponding equipment data, includes: Using 3D modeling tools, 3D physical models of corresponding process sub-processes are constructed based on the basic information of each processing equipment. Simulation software is then used to simulate the processing of the integrated pulping machine in different process sub-processes based on the 3D physical model of the integrated pulping machine to obtain its 3D simulation model. Data of one type of equipment in different process sub-processes of the integrated pulping machine is uploaded to a three-dimensional simulation model for synchronization. The values of various simulation parameters during the simulation process are adjusted. The simulation parameters refer to the electromagnetic parameters, thermodynamic parameters, and performance degradation parameters of the processing equipment in the actual application scenario. The temperature, content, and viscosity of the pulp in the three-dimensional simulation model under different values of simulation parameters are obtained. When the obtained temperature, content, and viscosity are the same as the data of the corresponding process sub-flow at the corresponding time, the three-dimensional simulation model at this time is used as the digital twin model of pulping, and the simulation parameters of the corresponding values are used as the digital twin parameters of the digital twin model.
[0018] It should be further explained that, in the specific implementation process, the process of obtaining the adjustable parameters contained in each process sub-process and obtaining the slurry viscosity under the dynamic changes of different adjustable parameters in the digital twin model includes: The adjustable parameters refer to the parameters that can be manually adjusted in each process sub-process, including starch feeding amount, water addition amount, type and amount of auxiliary materials, stirring speed, heating rate, gelatinization temperature and its holding time, and cooling rate. In the digital twin model, a single adjustable parameter (e.g., starch feeding amount) is taken as a control object, and other adjustable parameters are kept unchanged and taken as non-control objects. The final pulp consistency of the pulp in the integrated beater is obtained when the control object is at different values. The pulp consistency of the control object at different values and the values of the non-control objects are included in the same set, which is taken as the control object parameter set of the control object at different values. This method is used to obtain the control object parameter set of the corresponding control object at different values when other single adjustable parameters (e.g., water feeding amount) are taken as control objects and other adjustable parameters except them are taken as non-control objects and kept unchanged.
[0019] It should be further explained that, in the specific implementation process, the process of constructing a viscosity evaluation model according to different adjustable parameters and their corresponding pulp consistencies includes: According to the values of the corresponding control object and the non-control objects in the different control object parameter sets, and the final pulp consistency, a viscosity evaluation set is generated, which is divided into a training set and a test set; A convolutional neural network is constructed, and the values of different control objects and their non-control objects in the training set are taken as the input data of the convolutional neural network, and the corresponding pulp consistencies in the training set are taken as the output data of the convolutional neural network; The training set is used to train the convolutional neural network to obtain an initial convolutional neural network, and the test set is used to verify the model of the initial convolutional neural network. The initial convolutional neural network with an output less than or equal to a preset test error threshold is taken as the corresponding viscosity evaluation model.
[0020] It should be further explained that, in the specific implementation process, the process of obtaining a predicted consistency using the viscosity evaluation model, combining the preset target consistency to determine whether viscosity control is needed, and generating a control scheme includes: In actual application scenarios, the real-time values of the adjustable parameters in the integrated beater are obtained and input into the viscosity evaluation model to obtain the final pulp consistency according to the current values, which is taken as the predicted consistency. The target consistency is a preset value range. The obtained predicted consistency is compared with the preset target consistency. When the predicted consistency is within the target consistency range (including equal), no other operation is performed. When the predicted viscosity is not in the target viscosity range, it is determined that viscosity control is needed for the current process sub-process, any one of the controllable parameters is taken as the control object, and the other controllable parameters are taken as non-control objects, the viscosity evaluation model is used to obtain the predicted viscosity corresponding to different values of the control object when the non-control objects remain unchanged, and when the obtained predicted viscosity is in the target viscosity range, the value at this time is taken as the control parameter value; By taking this method, the control parameter value that can make the predicted viscosity in the target viscosity range is obtained when different controllable parameters are taken as the control object, and the total cost required for adjusting the different control objects from their current values to the corresponding control parameter values is obtained, including material cost and energy consumption cost, the control object with the lowest total cost and its control parameter value are taken as the corresponding control scheme, and it is fed back to the relevant personnel to prompt them to control the viscosity.
[0021] In the embodiments of the present application, the sweet potato vermicelli beating process optimization method based on viscosity dynamic control comprises the following steps: Step S1: obtaining the process whole flow of beating and dividing it into different process sub-processes, and obtaining the basic information and equipment data of the corresponding processing equipment of each process sub-process; Step S2: constructing a three-dimensional physical model according to the basic information of each processing equipment, and constructing a digital twin model of beating in combination with the corresponding equipment data; Step S3: obtaining the controllable parameters contained in each process sub-process, obtaining the slurry viscosity under the dynamic change of different controllable parameters in the digital twin model, and constructing a viscosity evaluation model according to different controllable parameters and their corresponding slurry viscosity; Step S4: obtaining the predicted viscosity by using the viscosity evaluation model, combining the preset target viscosity to determine whether viscosity control is needed, and generating a control scheme by using the viscosity evaluation model.
[0022] The above embodiments are only used to illustrate the technical method of the present application and are not limited. Although the present application has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical method of the present application can be modified or replaced equivalently without departing from the spirit and scope of the technical method of the present application.
Claims
1. A sweet potato vermicelli beating process optimization system based on dynamic viscosity control, characterized in that, The method comprises the following modules: a process division module, configured to obtain a whole process of pulping and divide the whole process into different process sub-processes, and obtain basic information and equipment data of processing equipment corresponding to each process sub-process; a model construction module, configured to construct a three-dimensional physical model of each processing equipment according to the basic information, and construct a digital twin model of pulping in combination with the corresponding equipment data; a viscosity evaluation module, configured to obtain adjustable parameters contained in each process sub-process, obtain pulp viscosity under dynamic changes of different adjustable parameters in the digital twin model respectively, and construct a viscosity evaluation model according to different adjustable parameters and corresponding pulp viscosities; a viscosity control module, configured to obtain a predicted viscosity by using the viscosity evaluation model, judge whether viscosity control is needed in combination with a preset target viscosity, and generate a control scheme by using the viscosity evaluation model.
2. The sweet potato vermicelli beating process optimization system based on dynamic viscosity regulation according to claim 1, characterized in that, The process of obtaining the whole process and dividing the whole process into different process sub-processes comprises: The whole process refers to the whole process of processing water, starch and auxiliary materials into pulp, which comprises the following process links: raw material preparation and processing, primary mixing and sizing, high-temperature gelatinization and curing, cooling and viscosity control. Different process links correspond to different processing equipment, and the processing process of the processing equipment corresponding to different process links is taken as a process sub-process.
3. The sweet potato vermicelli beating process optimization system based on dynamic viscosity regulation according to claim 2, characterized in that, The process of obtaining the basic information and equipment data of each process sub-process comprises: The basic information refers to specification parameters of the processing equipment corresponding to each process sub-process, and the equipment data comprises one type of equipment data and two types of equipment data. One type of acquisition unit is arranged for each processing equipment, and two types of acquisition unit are arranged for an integrated pulper. One type of equipment data of the processing equipment is acquired in real time by the one type of acquisition unit, which comprises pressure, temperature, load rate, current, voltage and vibration amplitude. Two types of equipment data of the integrated pulper are acquired in real time by the two types of acquisition unit, which refers to the temperature, content and viscosity of the pulp inside the integrated pulper.
4. The sweet potato vermicelli beating process optimization system based on dynamic viscosity regulation according to claim 3, characterized in that, The process of constructing the digital twin model of pulping comprises: A three-dimensional physical model of the corresponding process sub-process is constructed by using a three-dimensional modeling tool according to the basic information of each processing equipment, and a three-dimensional simulation model is obtained by simulating the processing process of the integrated pulper in different process sub-processes on the basis of the three-dimensional physical model of the integrated pulper by using simulation software. The one type of equipment data of the integrated pulper in different process sub-processes is uploaded to the three-dimensional simulation model for synchronization, the values of various simulation parameters in the simulation process are adjusted, and the temperature, content and viscosity of the pulp in the three-dimensional simulation model under different values of the simulation parameters are obtained. When the obtained temperature, content and viscosity are the same as the two types of equipment data of the corresponding process sub-process at the corresponding time, the three-dimensional simulation model at this time is taken as the digital twin model of pulping.
5. The sweet potato vermicelli beating process optimization system based on dynamic viscosity regulation according to claim 4, characterized in that, The process of obtaining the pulp viscosity under dynamic changes of different adjustable parameters comprises: The adjustable parameters comprise starch feeding amount, water feeding amount, auxiliary material type and amount, stirring speed, heating rate, gelatinization temperature and holding time, and cooling rate. In the digital twin model, the single adjustable parameter is taken as the control object, and other adjustable parameters are taken as non-control objects and remain unchanged. The final pulp consistency of the pulp in the integrated beater is obtained when the control object is at different values. The pulp consistency of the control object at different values and the values of the non-control objects are respectively included in the control object parameter set of the control object at different values. The control object parameter set corresponding to the control object at different values is obtained when other single adjustable parameters are taken as the control object and other adjustable parameters are taken as non-control objects and remain unchanged.
6. The sweet potato vermicelli beating process optimization system based on dynamic viscosity regulation according to claim 5, characterized in that, The process of constructing the viscosity evaluation model includes: According to the values of the corresponding control object and the non-control objects in the different control object parameter sets, and the final pulp consistency, a viscosity evaluation set is generated, and it is divided into a training set and a test set; A convolutional neural network is constructed, and the values of different control objects and their non-control objects in the training set are taken as the input data of the convolutional neural network, and the corresponding pulp consistency in the training set is taken as the output data of the convolutional neural network; The training set is used to train the convolutional neural network to obtain an initial convolutional neural network, and the test set is used to verify the model of the initial convolutional neural network. The initial convolutional neural network with an output less than or equal to a preset test error threshold is taken as the viscosity evaluation model.
7. The sweet potato vermicelli beating process optimization system based on dynamic viscosity regulation according to claim 6, characterized in that, The process of obtaining the predicted consistency and generating the control scheme includes: In the actual application scenario, the real-time values of the adjustable parameters in the integrated beater are obtained, and the viscosity evaluation model is input to obtain the corresponding predicted consistency. The predicted consistency is compared with the preset target consistency. When the predicted consistency is not within the target consistency range, it is determined that the current process sub-flow needs to be adjusted. Any adjustable parameter is taken as the control object, and other adjustable parameters are taken as non-control objects. The viscosity evaluation model is used to obtain the predicted consistency of the control object at different values when the non-control objects remain unchanged. When the predicted consistency is within the target consistency range, the value of the control object at this time is taken as the control parameter value. The control parameter values that can make the predicted consistency within the target consistency range are obtained when different adjustable parameters are taken as the control object. The total cost required to adjust the different control objects from their current values to the corresponding control parameter values is obtained. The control object and its control parameter value with the lowest total cost are taken as its control scheme, and it is fed back to the relevant personnel.
8. A method for optimizing the beating process of sweet potato vermicelli based on dynamic viscosity control, which is implemented based on the sweet potato vermicelli beating process optimization system of any one of claims 1-7, characterized in that, The method includes: Step S1: obtaining the process flow of beating and dividing it into different process sub-flows, and obtaining the basic information and equipment data of the corresponding processing equipment of each process sub-flow; Step S2: constructing a three-dimensional physical model according to the basic information of each processing equipment, and constructing a digital twin model of beating combined with the corresponding equipment data; Step S3: Obtain the adjustable parameters contained in each process sub-flow, obtain the slurry viscosity under the dynamic change of different adjustable parameters in the digital twin model respectively, and construct a viscosity evaluation model according to different adjustable parameters and the corresponding slurry viscosity; Step S4: Obtain the predicted viscosity by using the viscosity evaluation model, judge whether the viscosity needs to be adjusted according to the preset target viscosity, and generate an adjustment scheme by using the viscosity evaluation model.