A nanotube growth parameter optimization control method and system based on deep learning

Abstract

This invention discloses a method and system for optimizing and controlling nanotube growth parameters based on deep learning, relating to the field of nanotube production technology. It includes information acquisition: acquiring past nanotube growth information and user requirements; the past nanotube growth information includes past growth process information and nanotube performance information. This invention explores the differences between real-time environmental parameters and parameters in the target growth information to obtain the difference results. An analysis method is used to analyze the difference results to obtain the impact of representative parameter differences on the nanotubes, i.e., the difference consequences. An optimization method is then used to optimize and adjust the target growth information of the nanotubes based on the difference consequences to obtain optimized target growth information. This facilitates the production of nanotubes that meet user requirements. Furthermore, the optimization process reduces the probability of a large parameter space leading to slow search speed and getting trapped in local optima, thus improving the overall efficiency of the optimization.

Description

Technical Field

[0001] This invention relates to the field of nanotube production technology, specifically to a method and system for optimizing and controlling nanotube growth parameters based on deep learning. Background Technology

[0002] Nanotubes, as nanomaterials with unique physical and chemical properties, are widely used in electronic devices, sensors, composite materials and other fields. Their performance is highly dependent on the control of parameters during the growth process, so optimizing these parameters is crucial to the quality and performance of nanotubes.

[0003] A patent publication (CN119361030A) discloses an AI-assisted carbon nanotube growth control method and system. This system includes a three-dimensional AI virtual platform that collects carbon nanotube growth data and environmental data per unit time to construct a carbon nanotube model. The growth data includes growth rate, direction, and diameter, while the environmental data involves gas flow rate and electric field strength. The carbon nanotube model is mapped onto a three-dimensional coordinate system, where the X, Y, and Z axes represent the length, width, and height of the carbon nanotube, respectively, and the initial coordinates are recorded. This invention, through the use of a three-dimensional AI virtual platform and real-time data acquisition technology, can accurately capture environmental changes during carbon nanotube growth. By combining growth data with dynamic adjustments to environmental parameters during growth, it enables rapid response to environmental changes during actual control, avoiding the growth instability or deviations caused by the inability to adjust in a timely manner in traditional methods.

[0004] In existing technologies, the optimization and control of nanotube growth parameters often involves adjusting multiple parameters to obtain optimal performance. The adjustment process employs incremental experiments and response surface methodology. However, this approach suffers from drawbacks such as a large parameter space leading to slow search speed and the tendency to get trapped in local optima, resulting in limited overall efficiency and effectiveness. This hinders the optimization of nanotube growth based on actual demand information. Furthermore, errors caused by adjusting parameters in production machines cannot be addressed in a timely manner, making it difficult to produce nanotubes with performance corresponding to user requirements. Therefore, this invention is proposed. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for optimizing and controlling nanotube growth parameters based on deep learning, so as to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for optimizing and controlling nanotube growth parameters based on deep learning, comprising:

[0007] Information Acquisition: Acquire past nanotube growth information and user requirements. Past nanotube growth information includes past growth process information and nanotube performance information.

[0008] Information processing: Establish an information database to store past nanotube growth information, establish the correlation between past growth process information and nanotube performance information, and recommend target nanotube growth information to users based on user needs and in conjunction with the information database;

[0009] include:

[0010] Growth information determination: Determine whether the user needs to generate nanotube growth information and obtain the determination result. When the determination result is that the user needs to generate nanotube growth information, the target growth information is obtained by generating nanotube growth information that meets the user's needs based on the user's needs and the information database through the demand matching method.

[0011] Monitoring status: Monitor the real-time environmental parameters of nanotube growth at a unified time point. Obtain the difference results between the real-time environmental parameters and the parameters in the target growth information based on the same time point. Based on the results, obtain the consequences of the difference through analysis methods.

[0012] Parameter optimization: Based on the difference consequences, the subsequent parameters in the target growth information are optimized and controlled by the optimization method to obtain the optimized target growth information;

[0013] Information output: Nanotube growth based on optimized target growth information;

[0014] Feedback Update: Based on the optimized target growth information, the information in the information database is supplemented by an update method to obtain an updated information database.

[0015] Furthermore, the process of obtaining difference information based on the difference between real-time environmental parameters and target growth information obtained at the same time point is as follows: split the real-time environmental parameters to obtain several first parameter items, split the target growth information to obtain several second parameter items, compare the first parameter items and the second parameter items to obtain comparison results, and extract the second parameter items that differ from the first parameter items and the magnitude of the difference to obtain difference information.

[0016] Furthermore, the analysis method includes: splitting the difference results to obtain difference items and difference magnitudes; exploring the performance impact caused by differences in different growth parameters based on an information database using a relationship acquisition method to obtain consequence information, the consequence information including difference information and impact information corresponding to the difference information; performing information matching in the difference information based on the difference items to obtain matching results; extracting the corresponding impact information based on the matching results to obtain target impact information; and judging the specific details of the target impact information based on the difference magnitude and the target impact information using a depth acquisition method to obtain specific impact information, the specific impact information being the difference consequences.

[0017] Furthermore, the relationship acquisition method includes: pre-setting the number of identical parameter parts to obtain the processing quantity; extracting past growth process information with the same corresponding parameters from the past growth process information in the information database to obtain several growth information to be processed; classifying the several growth information to be processed based on the processing quantity to obtain the target category and the processing growth information corresponding to the target category; extracting the nanotube performance information of the processing growth information from the information database based on correlation to obtain the processing performance; extracting the processing growth information in the target category to obtain the reference growth information; extracting the nanotube performance information of the reference growth information to obtain the reference performance; obtaining the differences between several reference growth information to obtain the difference information; obtaining the differences between the processing performance of the reference growth information that generated the difference information to obtain the impact information; and integrating the difference information and the impact information to obtain the consequence information.

[0018] Furthermore, the depth acquisition method includes: splitting the target influence information to obtain several sub-influence information, determining the corresponding previous nanotube growth information based on the sub-influence information to obtain two boundary growth information, obtaining the parameter difference in the two boundary growth information to obtain the target size, integrating all target sizes to obtain a target set, traversing the target set by the difference size to find the target size closest to the difference size to obtain the selected size, and determining the sub-influence information based on the selected size to obtain specific influence information.

[0019] Furthermore, the optimization method includes: determining solution information based on the consequences of differences, determining real-time growth progress, removing real-time production progress from the target growth information to obtain subsequent growth information, determining whether there are adjustable parameters in the subsequent growth information that are associated with the solution information to obtain an optimization result, when the optimization result indicates that there are adjustable parameters in the subsequent growth information that are associated with the solution information, extracting the adjustable parameters in the subsequent growth information that are associated with the solution information to obtain target parameters, optimizing and adjusting the target parameters based on the impact information in conjunction with the target parameters to obtain optimized parameters, adjusting the target parameters in the subsequent growth information to the corresponding optimized parameters to obtain optimized target growth information, and when the optimization result indicates that there are no adjustable parameters in the subsequent growth information that are associated with the solution information, extracting the solution information and real-time growth progress to generate feedback information.

[0020] Furthermore, the updating method includes: recording the optimized target growth information and the performance of the nanotubes produced by the optimized target growth information to obtain real-time nanotube growth information, storing the real-time nanotube growth information in an information database, recording the optimization process of the target growth information to obtain the optimization process, and storing the optimization process in an information database.

[0021] Furthermore, the demand matching method includes: determining the target performance based on user demand; traversing the information database based on the target performance to obtain traversal results; when the traversal results indicate that nanotube performance information corresponding to the target performance exists in the information database, extracting previous nanotube growth information corresponding to the target performance from the information database based on correlation to obtain target growth information; when the traversal results indicate that nanotube performance information corresponding to the target performance does not exist in the information database, extracting nanotube performance information exceeding the target performance from the information database based on correlation to obtain selected nanotube performance information; and extracting previous nanotube growth information corresponding to the selected nanotube performance information from the information database based on correlation to obtain target growth information.

[0022] A deep learning-based nanotube growth parameter optimization control system is provided, which utilizes the aforementioned deep learning-based nanotube growth parameter optimization control method.

[0023] Compared with the prior art, the beneficial effects of the present invention are:

[0024] This deep learning-based nanotube growth parameter optimization control method and system explores the differences between real-time environmental parameters and target growth information to obtain the difference results. An analysis method is then used to analyze these differences to identify the impact of these parameter differences on the nanotubes, i.e., the difference consequences. An optimization method is then used to adjust the target growth information of the nanotubes based on these difference consequences, resulting in optimized target growth information. This optimizes the production of nanotubes that meet user requirements. Furthermore, the optimization process reduces the likelihood of a large parameter space leading to slow search speed and getting trapped in local optima, thus improving overall optimization efficiency. By combining actual demand information and real-time environmental information, the nanotube growth parameter adjustment and optimization are tailored to user needs.

[0025] Simultaneously, by analyzing the differences between real-time environmental parameters and target growth information, the errors of specific parameter equipment can be identified. This allows staff to further determine whether the parameter equipment has been damaged. Through a set demand-matching method, the system searches the information database for nanotube growth information corresponding to the user's needs to obtain target growth information. This allows for the recommendation of suitable growth processes and parameters when the user only has nanotube performance requirements. When the user has their own nanotube growth information, it extracts that information to obtain target growth information. Through a set update method, the optimized target growth information and the performance information of the produced nanotubes are collected in real time and added to the information database to expand its resources and enrich its information content, facilitating more reasonable optimization and analysis.

[0026] Meanwhile, by using the set depth acquisition method, it is possible to determine the corresponding values ​​of specific performance items affected by the differing parameter items, so as to facilitate better optimization. Under the premise of optimization, the optimization method optimizes the growth information of nanotubes to make the growth of nanotubes meet the user's needs. Under the premise of non-optimization, feedback information is generated, which can be fed back to the user through the user's address information, so that the user can understand the specific situation of nanotube production and optimize the subsequent growth of nanotubes based on the non-optimization information, thereby improving the efficiency of nanotube production and comprehensively optimizing nanotube production. Attached Figure Description

[0027] Figure 1 This is a schematic diagram of the overall process structure of the present invention;

[0028] Figure 2 This is a schematic diagram of the analytical method of the present invention;

[0029] Figure 3 This is a schematic diagram of the structure of Embodiment 2 of the present invention;

[0030] Figure 4 This is a schematic diagram of the depth acquisition method of the present invention. Detailed Implementation

[0031] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0032] Different applications have different performance requirements for nanotubes. For example, electronic devices require high conductivity, while composite materials require good mechanical properties. By optimizing parameters such as tube diameter, number of tube wall layers, and axial alignment, the electronic band structure and conductivity of nanotubes can be controlled. For example, small-diameter carbon nanotubes have higher electron mobility, while axially aligned nanotubes exhibit excellent conductivity. Optimizing process parameters (such as low-temperature growth technology) can reduce energy consumption and defect density. For example, microwave plasma CVD can achieve efficient growth at around 600℃, while traditional methods require higher temperatures.

[0033] like Figures 1-4 As shown, the present invention provides a technical solution: a method for optimizing and controlling nanotube growth parameters based on deep learning, comprising:

[0034] Information Acquisition: Acquire past nanotube growth information and user requirements. Past nanotube growth information includes past growth process information and nanotube performance information.

[0035] It is important to note that the previous growth process information contains specific parameters and procedures for the production of nanotubes. These parameters may include temperature, gas pressure, gas composition, atmosphere flow rate, catalyst concentration, etc. The nanotube performance information contains the properties of the nanotubes, which may include mechanical, electrical, thermal, and optical properties, etc. The user requirements information contains the user's specific needs for producing nanotubes.

[0036] Information processing: Establish an information database to store past nanotube growth information, establish the correlation between past growth process information and nanotube performance information, and recommend target nanotube growth information to users based on user needs and in conjunction with the information database;

[0037] It is important to note that by establishing a database to store past nanotube growth information, we can provide a storage space for past nanotube growth information, which will facilitate subsequent information matching and search operations. At the same time, by establishing correlations, we can easily find nanotube performance information and past growth process information by matching past growth process information or nanotube performance information.

[0038] Its features include:

[0039] Growth information determination: Determine whether the user needs to generate nanotube growth information and obtain the determination result. When the determination result is that the user needs to generate nanotube growth information, the target growth information is obtained by generating nanotube growth information that meets the user's needs based on the user's needs and the information database through the demand matching method.

[0040] It is important to note that by setting a demand matching method, the system searches the information database for nanotube growth information corresponding to the user's needs to obtain the target growth information. This is so that when the user only has nanotube performance requirements, the system can recommend a suitable growth process and parameters to the user. When the user has their own nanotube growth information, the system can extract the user's own nanotube growth information to obtain the target growth information.

[0041] Monitoring status: Monitor the real-time environmental parameters of nanotube growth at a unified time point. Based on the same time point, obtain the difference between the real-time environmental parameters and the parameters in the target growth information to obtain the difference results. Based on the difference results, obtain the difference consequences through analysis methods.

[0042] It is important to note that the real-time environmental parameters for nanotube growth include multiple parameters. The specific sub-items of the real-time environmental parameters correspond to the parameters during the nanotube growth process. Based on the same point in time, this involves aligning the environmental parameters of the real-time growing nanotubes with the specific growth parameters. After alignment, deviations in the environmental parameters are identified to obtain the discrepancies. The discrepancies are then analyzed using the set analysis methods to determine the consequences of these discrepancies. Simultaneously, this stage allows for comparison between the real-time environmental parameters and the nanotube growth process information to determine if there are any deviations in the specific parameter equipment, enabling staff to adjust the parameters and equipment in a timely manner.

[0043] Parameter optimization: Based on the difference consequences, the subsequent parameters in the target growth information are optimized and controlled by the optimization method to obtain the optimized target growth information;

[0044] It is important to note that, based on the consequences of the differences, the subsequent parameters in the target growth information are optimized and controlled through the set optimization methods to obtain optimized target growth information, so as to ensure that the performance information of the produced nanotubes is consistent with the user's requirements.

[0045] Information output: Nanotube growth based on optimized target growth information;

[0046] Feedback Update: Based on the optimized target growth information, the information in the information database is supplemented by an update method to obtain an updated information database.

[0047] It is important to note that by using the set update method, the optimized target growth information and the performance information of the produced nanotubes are collected in real time and added to the information database to expand the resources in the database and enrich the amount of information, so as to make more reasonable optimizations and analyses.

[0048] Example 1:

[0049] In a nanotube production line requiring optimized control of nanotube growth parameters, chemical vapor deposition is used to produce nanotubes. Past nanotube growth information can be obtained by collecting historical production information from the nanotube production line. An information receiving interface is established to receive user demand information. An information repository is built to store past nanotube growth information. Before storage, the correlation between past growth process information and nanotube performance information is established based on the production system of the nanotube production line. During the method's operation, it is determined whether the demand information in the information receiving interface is the target growth information that needs to be recommended. If not, the target growth information is directly obtained based on the production information provided by the user. If it is required, the target growth information is obtained by searching the information repository for growth information corresponding to the user's demand, in accordance with the user's requirements.

[0050] The process of obtaining difference information based on the difference between real-time environmental parameters and target growth information obtained at the same time point is as follows: split the real-time environmental parameters to obtain several first parameter items, split the target growth information to obtain several second parameter items, compare the first parameter items and the second parameter items to obtain the comparison results, and extract the second parameter items that differ from the first parameter items and the magnitude of the difference to obtain the difference information.

[0051] It should be noted that the difference information specifically refers to the parameter differences between the two at the same point in time. For example, the parameter temperature needs to be set to 750°C, while the measured real-time environmental parameter is that the reaction chamber temperature is 742°C. In this case, the parameter that differs between the two is temperature, and the difference is 8°C.

[0052] like Figure 2 As shown, the analysis method includes: splitting the difference results to obtain difference items and difference magnitudes; exploring the performance impact caused by differences in different growth parameters through relation acquisition methods based on the information database to obtain consequence information, which includes difference information and impact information corresponding to the difference information; performing information matching in the difference information based on the difference items to obtain matching results; extracting the corresponding impact information based on the matching results to obtain target impact information; and judging the specific details of the target impact information based on the difference magnitude and the target impact information through depth acquisition methods to obtain specific impact information, which is the difference consequence.

[0053] It is important to note that the relationship acquisition method used to explore the performance impact caused by differences in different growth parameters yields consequence information. Specifically, this can be understood as acquiring information on several previous growth processes with differences in one or more parameters, determining the differences in nanotube performance information, and thus obtaining the impact of specific parameters on specific performance items in the nanotube performance information. By acquiring consequence information, it is convenient to obtain specific detailed impact information through the set depth acquisition method, so as to better optimize the method. By using the set depth acquisition method, combined with the specific difference magnitude, specific impact details are obtained to obtain specific impact information. This can be understood as obtaining the specific performance items affected by the differing parameter items through the first half of the analysis method, and then determining the corresponding values ​​of the specific performance items affected by the differing parameter items through the set depth acquisition method, so as to carry out better optimization.

[0054] like Figure 2As shown, the relationship acquisition method includes: pre-setting the number of identical parameters to obtain the processing quantity; extracting past growth process information with the same parameters from the information database to obtain several growth information to be processed; classifying the several growth information to be processed based on the processing quantity to obtain the target category and the processing growth information corresponding to the target category; extracting the nanotube performance information of the processing growth information from the information database based on correlation to obtain the processing performance; extracting the processing growth information from the target category to obtain the reference growth information; extracting the nanotube performance information of the reference growth information to obtain the reference performance; obtaining the differences between several reference growth information to obtain the difference information; obtaining the differences between the processing performance of the reference growth information that generated the difference information to obtain the impact information; and integrating the difference information and the impact information to obtain the consequence information.

[0055] Example 2:

[0056] The processing quantities are 1, 2, 3, 4, 5, 6, and 7. Any previous growth process information from the information database is selected as the original matching object. The original matching object is traversed through the information database to find previous growth process information with the same parameter part and the same parameter value to obtain the growth information to be processed. When the growth information to be processed contains only one identical parameter part with the same specific value, it is considered a target analogy with a processing quantity of 1. This process continues, and so on. Figure 3 As shown, Figure 3 In the diagram, 'a' represents the classification of the growth information to be processed into a target category with a processing quantity of 1. Figure 3 In the diagram, 'b' represents the classification of the growth information to be processed into a target category with a processing quantity of 2. Figure 3 The squares above and below represent the same parameter, and their shaded parts are the specific values. The shaded parts above and below each other have the same value.

[0057] It is important to note that the processing quantity refers to a specific number of parameters, including single or multiple parameters. The specific processing quantity can be determined based on the number of parameters, such as 1, 2, 3, 4, etc. The process involves extracting past nanotube growth information with the same parameters from the information database to obtain several growth information to be processed. That is, selecting past nanotube growth information with the same parameters and the same specific value of the same parameter to obtain the growth information to be processed. The growth information to be processed is classified according to the processing quantity. Classification facilitates multi-threaded processing of data information, thereby improving the efficiency of information processing. Specifically, the process involves classifying past nanotube growth information with the same parameters and the same specific value of the same parameter according to the number of the same parameters. The number of specific target categories is consistent with the processing quantity. By obtaining the differences between the growth information and the differences in nanotube performance, the impact information reflecting the differences can be obtained.

[0058] like Figure 4 As shown, the depth acquisition method includes: splitting the target influence information to obtain several sub-influence information, determining the corresponding previous nanotube growth information based on the sub-influence information to obtain two boundary growth information, obtaining the parameter difference in the two boundary growth information to obtain the target size, integrating all target sizes to obtain the target set, traversing the target set by the difference size to find the target size closest to the difference size to obtain the selected size, and determining the sub-influence information based on the selected size to obtain the specific influence information.

[0059] It is important to note that the number of target influence information can be single or multiple. Based on the sub-influence information, the nanotube growth information that generates the sub-influence information is determined to obtain two boundary growth information. The sub-influence information is the result of a performance comparison between the two growth information. The target size is obtained by obtaining the parameter difference between the two boundary growth information. The target size here includes the specific name and specific value of its parameters. The target set includes multiple target sizes, and may also include a single target size. The process of obtaining specific influence information based on the selected size and determining the sub-influence information is that the target size closest to the difference size is selected from the target set to obtain the selected size. By determining the selected size, the specific influence information can be estimated, that is, the specific difference between the subsequent growth of the nanotube and the performance required by the user.

[0060] like Figure 1 As shown, the optimization method includes: determining the solution information based on the consequences of the difference, determining the real-time growth progress, removing the real-time production progress from the target growth information to obtain subsequent growth information, determining whether there are adjustable parameters in the subsequent growth information that are associated with the solution information to obtain the optimization result, when the optimization result indicates that there are adjustable parameters in the subsequent growth information that are associated with the solution information, extracting the adjustable parameters in the subsequent growth information that are associated with the solution information to obtain the target parameter, optimizing and adjusting the target parameter based on the impact information and the target parameter to obtain the optimized parameter, adjusting the target parameter in the subsequent growth information to the corresponding optimized parameter to obtain the optimized target growth information, and when the optimization result indicates that there are no adjustable parameters in the subsequent growth information that are associated with the solution information, extracting the solution information and the real-time growth progress to generate feedback information.

[0061] Example 3:

[0062] A differential consequence identification model is established. This model outputs information to address specific differential consequences. Specifically, it collects performance item differences and parameter differences to obtain training data, labels the parameter differences in the training data, selects a deep learning model as the model base, imports the training data into the model base for training, and obtains the initial model. Validation information is then determined, including validation performance item differences and validation parameter differences. The validation performance item differences are imported into the deep learning model to obtain the derived results, which are compared with the validation parameters to obtain the comparison results. Based on the comparison results, the initial model is optimized and adjusted to obtain the differential consequence identification model.

[0063] It is important to note that the process of determining solution information based on differential consequences involves obtaining solution information by deriving parameters associated with the specific consequences identified in the differential consequences. Specifically, this can be achieved by training a dedicated model to determine the solution information based on the differential consequences. Through a set optimization method, the growth information of nanotubes is optimized under optimizable conditions to ensure that the growth of nanotubes meets user needs. Under non-optimizable conditions, feedback information is generated, specifically through the user's address information, to allow the user to understand the specific situation of nanotube production. Based on the non-optimizable information, subsequent nanotube growth can be optimized to improve the efficiency of nanotube production and comprehensively optimize nanotube production.

[0064] Example 4:

[0065] The influence of specific parameters on performance was investigated using a relational acquisition method. Increasing temperature improved the deposition rate and diameter; increasing NH3 flow rate improved nickel nucleation and increased diameter; adjusting pressure altered the gas diffusion rate, affecting the deposition rate. The measured real-time environmental parameters were as follows: temperature: 700°C, gas flow rate: O2 50 sccm, NH3 30 sccm, pressure: 5 Torr; real-time diameter: 45 nm (deviation 5 nm); real-time length: 8 μm (deviation 2 μm); nucleation rate: 2 nuclei per second; growth rate: 0.8 μm / h. The parameters in the target growth information were adjusted as follows: temperature: 730°C, gas flow rate: 50 sccm, NH3 30 sccm, pressure: 5 Torr; real-time diameter: 45 nm (deviation 5 nm); real-time length: 8 μm (deviation 2 μm); nucleation rate: 2 nuclei per second; growth rate: 0.8 μm / h. The parameters are: O2 50 sccm, NH3 40 sccm, and pressure 5 Torr. The differences between these parameters and those in the growth information are temperature and gas flow rate. Based on these differences, the influence of temperature and gas flow rate is determined. The influence of temperature and gas flow rate is on the growth rate and diameter. The subsequent optimization process aims to increase the temperature (suggested range: 700-750°C), adjust the gas flow rate (increase the NH3 flow rate to 40 sccm), or adjust the pressure. Increasing the temperature and adjusting the gas flow rate are the optimized parameters, i.e., the optimized target growth information. The optimized target growth information is used to increase the diameter and growth rate of the nanotubes.

[0066] like Figure 1 As shown, the update method includes: recording the optimized target growth information and the performance of the nanotubes produced by the optimized target growth information to obtain real-time nanotube growth information, storing the real-time nanotube growth information in an information database, recording the optimization process of the target growth information to obtain the optimization process, and storing the optimization process in an information database.

[0067] It is important to note that by recording the optimized target growth information and the optimized nanotube performance, real-time nanotube growth information can be obtained. Storing this real-time nanotube growth information in the information database can improve the richness of the information database and achieve continuous learning. Adding different parameters with the same differential consequences to the influence information can improve the accuracy of subsequent optimization and determination of influence information. By recording the optimization process in the information database, corresponding matching can be performed directly in the future. When the same target growth information and the same differential information are available, the optimization process can be directly output for optimization.

[0068] like Figure 1 As shown, the demand matching method includes: determining the target performance based on user needs; traversing the information database based on the target performance to obtain the traversal results; when the traversal results indicate that there is nanotube performance information in the information database corresponding to the target performance, extracting the previous nanotube growth information corresponding to the target performance from the information database based on correlation to obtain the target growth information; when the traversal results indicate that there is no nanotube performance information corresponding to the target performance in the information database, extracting nanotube performance information exceeding the target performance from the information database based on correlation to obtain the selected nanotube performance information; and extracting the previous nanotube growth information corresponding to the selected nanotube performance information from the information database based on correlation to obtain the target growth information.

[0069] It is important to note that by setting a demand matching method, the performance information of the nanotubes to be produced is determined based on the user's requirements to obtain the target performance. The information database is then traversed based on the target performance to obtain the traversal results. Based on the traversal results, it is determined whether the corresponding target growth information can be generated according to the user's requirements. At the same time, the process of generating target growth information can also be determined through a dedicated model. By inputting the nanotube performance requirements, the corresponding nanotube growth information is output. By setting the demand matching method, the efficiency of users in determining target growth information can be improved.

[0070] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended embodiments and their equivalents.

Claims

1. A method for optimizing and controlling nanotube growth parameters based on deep learning, comprising: Information Acquisition: Acquire past nanotube growth information and user requirements. Past nanotube growth information includes past growth process information and nanotube performance information. Information processing: Establish an information database to store past nanotube growth information, establish the correlation between past growth process information and nanotube performance information, and recommend target nanotube growth information to users based on user needs and in conjunction with the information database; Its features include: Growth information determination: Determine whether the user needs to generate nanotube growth information and obtain the determination result. When the determination result indicates that the user needs to generate nanotube growth information, the target growth information is obtained by generating nanotube growth information that meets the user's needs based on the user's needs and the information database through the demand matching method. Monitoring status: Monitor the real-time environmental parameters of nanotube growth at a unified time point. Based on the same time point, obtain the difference between the real-time environmental parameters and the parameters in the target growth information to obtain the difference results. Based on the results, obtain the difference consequences through analysis methods. Parameter optimization: Based on the difference consequences, the subsequent parameters in the target growth information are optimized and controlled by the optimization method to obtain the optimized target growth information; Information output: Nanotube growth based on optimized target growth information; Feedback Update: Based on the optimized target growth information, the information in the information database is supplemented by an update method to obtain an updated information database; The analysis method includes: splitting the difference results to obtain difference items and difference magnitudes; exploring the performance impact caused by differences in different growth parameters through a relation acquisition method based on an information database to obtain consequence information, which includes difference information and impact information corresponding to the difference information; performing information matching in the difference information based on the difference items to obtain matching results; extracting the corresponding impact information based on the matching results to obtain target impact information; and judging the specific details of the target impact information based on the difference magnitude and the target impact information through a depth acquisition method to obtain specific impact information, which is the difference consequence. The optimization method includes: determining solution information based on the consequences of differences, determining the real-time growth progress, removing the real-time production progress from the target growth information to obtain subsequent growth information, determining whether there are adjustable parameters in the subsequent growth information that are associated with the solution information to obtain an optimization result, when the optimization result indicates that there are adjustable parameters in the subsequent growth information that are associated with the solution information, extracting the adjustable parameters in the subsequent growth information that are associated with the solution information to obtain the target parameters, optimizing and adjusting the target parameters based on the impact information and the target parameters to obtain the optimized parameters, adjusting the target parameters in the subsequent growth information to the corresponding optimized parameters to obtain the optimized target growth information, and when the optimization result indicates that there are no adjustable parameters in the subsequent growth information that are associated with the solution information, extracting the solution information and the real-time growth progress to generate feedback information.

2. The method for optimizing and controlling nanotube growth parameters based on deep learning according to claim 1, characterized in that: The process of obtaining difference information based on the difference between real-time environmental parameters and target growth information obtained at the same time point is as follows: split the real-time environmental parameters to obtain several first parameter items, split the target growth information to obtain several second parameter items, compare the first parameter items and the second parameter items to obtain the comparison results, and extract the second parameter items that differ from the first parameter items and the magnitude of the difference to obtain the difference information.

3. The method for optimizing and controlling nanotube growth parameters based on deep learning according to claim 1, characterized in that: The relationship acquisition method includes: pre-setting the number of identical parameters to obtain the processing quantity; extracting past growth process information with the same parameters from the information database to obtain several growth information to be processed; classifying the several growth information to be processed based on the processing quantity to obtain the target category and the processing growth information corresponding to the target category; extracting the nanotube performance information of the processing growth information from the information database based on correlation to obtain the processing performance; extracting the processing growth information from the target category to obtain the reference growth information; extracting the nanotube performance information of the reference growth information to obtain the reference performance; obtaining the differences between several reference growth information to obtain the difference information; obtaining the differences between the processing performance of the reference growth information that generated the difference information to obtain the impact information; and integrating the difference information and the impact information to obtain the consequence information.

4. The method for optimizing and controlling nanotube growth parameters based on deep learning according to claim 1, characterized in that: The depth acquisition method includes: splitting the target influence information to obtain several sub-influence information; determining the corresponding previous nanotube growth information based on the sub-influence information to obtain two boundary growth information; obtaining the parameter difference in the two boundary growth information to obtain the target size; integrating all target sizes to obtain a target set; traversing the target set by the difference size to find the target size closest to the difference size to obtain the selected size; and determining the sub-influence information based on the selected size to obtain the specific influence information.

5. The method for optimizing and controlling nanotube growth parameters based on deep learning according to claim 1, characterized in that: The updating method includes: recording the optimized target growth information and the performance of the nanotubes produced by the optimized target growth information to obtain real-time nanotube growth information, storing the real-time nanotube growth information in an information database, recording the optimization process of the target growth information to obtain the optimization process, and storing the optimization process in an information database.

6. The method for optimizing and controlling nanotube growth parameters based on deep learning according to claim 1, characterized in that: The demand matching method includes: determining the target performance based on user needs; traversing the information database based on the target performance to obtain traversal results; when the traversal results indicate that nanotube performance information corresponding to the target performance exists in the information database, extracting past nanotube growth information corresponding to the target performance from the information database based on correlation to obtain target growth information; when the traversal results indicate that nanotube performance information corresponding to the target performance does not exist in the information database, extracting nanotube performance information exceeding the target performance from the information database based on correlation to obtain selected nanotube performance information; and extracting past nanotube growth information corresponding to the selected nanotube performance information from the information database based on correlation to obtain target growth information.

7. A deep learning-based nanotube growth parameter optimization control system, characterized in that: The method for optimizing and controlling nanotube growth parameters based on deep learning, as described in any one of claims 1-6, was used.

Images