Prediction and design of carbon dot-based long afterglow materials using machine learning

Abstract

The invention discloses a method for predicting and designing a carbon dot-based long afterglow material by utilizing machine learning. The method comprises the following specific steps: S1, collecting and sorting data; s2, feature selection and processing; s3, machine learning modeling: model selection, model training and model evaluation; s4, predicting and optimizing; and S5, verification and experiment: verifying the model and adjusting the experiment. The method has high efficiency, the optimal component combination can be rapidly screened out, and the number of experiments is reduced; the performance of the material can be optimized, and better luminescence performance can be realized by accurately adjusting components; the cost can be saved, and the material research and development time and economic cost are reduced; moreover, accurate prediction can be realized, the prediction accuracy and reliability are improved, and the trial and error processes are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of luminescent material manufacturing, and in particular to the use of machine learning to predict and design carbon dot-based long afterglow materials. Background Art

[0002] Carbon dot-based long afterglow materials are materials with long-lasting luminescence properties, usually with carbon dots (carbon quantum dots) as the matrix. These materials combine the excellent optical properties of carbon dots with the characteristics of long afterglow, and can continue to emit light for a long time after absorbing light sources. They are widely used in biological imaging, sensors, and display technologies. Carbon dot-based materials have attracted attention due to their excellent optical properties, biocompatibility, and high stability.

[0003] Existing methods for predicting and designing carbon dot-based long afterglow materials are inefficient and require a large number of experiments to determine the optimal material composition and conditions; they are costly, and the experimental process consumes time and resources, increasing R&D costs; the discovery cycle is long and lacks a systematic approach, which may extend the time it takes to discover new materials; there are many trial and errors, and multiple adjustments may be required during the experiment, increasing the cost of trial and error. Summary of the invention

[0004] The purpose of the present invention is to solve the shortcomings of the prior art and to propose the use of machine learning to predict and design carbon dot-based long afterglow materials.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] Using machine learning to predict and design carbon dot-based long afterglow materials includes the following specific steps:

[0007] S1: Data collection and organization:

[0008] S1a: Experimental data: Collect the experimental data of existing carbon dot materials;

[0009] S1b: Literature data: integrate the property data of carbon dot-based materials reported in relevant literature;

[0010] S2: Feature selection and processing:

[0011] S2a: Feature selection: extract key features that affect the long afterglow characteristics from the collected data;

[0012] S2b: Data preprocessing: standardize and preprocess the data, clean and organize the data, remove noise and incomplete information, so as to facilitate machine learning modeling and split the data into training and test sets;

[0013] S3: Machine Learning Modeling:

[0014] S3a: Model selection: Selecting a suitable machine learning algorithm;

[0015] S3b: Model training: Use the data from the training set to train the model so that it can learn the relationship between the properties of carbon dot-based long afterglow materials and their composition;

[0016] S3c: Model evaluation: Use the data from the test set to evaluate the predictive ability of the model and use appropriate evaluation metrics for validation;

[0017] S4: Prediction and Optimization:

[0018] S4a: Prediction: Use the trained model to predict the performance of new carbon dot-based long afterglow materials, determine which components and conditions may lead to excellent long afterglow effects, explore different combinations of material components and structures, and optimize the long afterglow characteristics;

[0019] S4b: Optimization: Optimization algorithms are used to further adjust the composition and preparation conditions to design carbon dot materials with the best performance;

[0020] S5: Verification and Experimentation:

[0021] S5a: Model verification: Apply the results predicted by the model to actual experiments, synthesize and characterize the predicted carbon dot-based long afterglow materials, and verify the accuracy of its predictions;

[0022] S5b: Experimental adjustment: Adjust the model and design scheme according to the experimental results, perform iterative optimization, gradually optimize the performance of the material, and improve prediction accuracy and design efficiency.

[0023] As a further technical solution of the present invention, in S1a, the experimental data includes composition, structure, performance, preparation conditions and optical properties.

[0024] As a further technical solution of the present invention, the optical characteristics include luminous intensity and duration.

[0025] As a further technical solution of the present invention, in S2a, key features include the size of carbon dots, surface functional groups, doping elements, and synthesis method parameters.

[0026] As a further technical solution of the present invention, in S3a, the machine learning algorithm is such as a regression model, a classification model or a deep learning model.

[0027] As a further technical solution of the present invention, regression models such as linear regression and support vector regression, classification models such as decision trees and random forests, and deep learning models such as neural networks.

[0028] As a further technical solution of the present invention, in S3c, the evaluation indicators are such as mean square error and accuracy.

[0029] As a further technical solution of the present invention, in S4b, the optimization algorithm is such as genetic algorithm or particle swarm optimization.

[0030] The beneficial effects of the present invention are:

[0031] This method is highly efficient in predicting and designing carbon dot-based long afterglow materials, can quickly screen out the best combination of ingredients and reduce the number of experiments; can optimize the performance of the material and achieve better luminescence performance by precisely adjusting the ingredients; can save costs and reduce the time and economic cost of material research and development; and can achieve precise predictions, improve the accuracy and reliability of the predictions, and reduce the trial and error process. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 A flow chart of the present invention for predicting and designing carbon dot-based long afterglow materials using machine learning. DETAILED DESCRIPTION

[0033] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0034] Please see attached Figure 1 , using machine learning to predict and design carbon dot-based long afterglow materials, including the following specific steps:

[0035] S1: Data collection and organization:

[0036] S1a: Experimental data: Collect the experimental data of existing carbon dot materials, including composition, structure, performance, preparation conditions and optical properties. The optical properties include luminescence intensity and duration.

[0037] S1b: Literature data: integrate the property data of carbon dot-based materials reported in relevant literature;

[0038] S2: Feature selection and processing:

[0039] S2a: Feature selection: extract key features that affect the long afterglow characteristics from the collected data, such as the size of carbon dots, surface functional groups, doping elements, and synthesis method parameters;

[0040] S2b: Data preprocessing: standardize and preprocess the data, clean and organize the data, remove noise and incomplete information, so as to facilitate machine learning modeling and split the data into training and test sets;

[0041] S3: Machine Learning Modeling:

[0042] S3a: Model selection: Select a suitable machine learning algorithm, such as regression model, classification model or deep learning model. Regression models include linear regression and support vector regression, classification models include decision tree and random forest, and deep learning models include neural network.

[0043] S3b: Model training: Use the data from the training set to train the model so that it can learn the relationship between the properties of carbon dot-based long afterglow materials and their composition;

[0044] S3c: Model evaluation: Use the data from the test set to evaluate the predictive ability of the model and use appropriate evaluation indicators for verification, such as mean square error and accuracy;

[0045] S4: Prediction and Optimization:

[0046] S4a: Prediction: Use the trained model to predict the performance of new carbon dot-based long afterglow materials, determine which components and conditions may lead to excellent long afterglow effects, explore different combinations of material components and structures, and optimize the long afterglow characteristics;

[0047] S4b: Optimization: Further adjust the composition and preparation conditions using optimization algorithms, such as genetic algorithm and particle swarm optimization, to design carbon dot materials with the best performance;

[0048] S5: Verification and Experimentation:

[0049] S5a: Model verification: Apply the results predicted by the model to actual experiments, synthesize and characterize the predicted carbon dot-based long afterglow materials, and verify the accuracy of its predictions;

[0050] S5b: Experimental adjustment: Adjust the model and design scheme according to the experimental results, perform iterative optimization, gradually optimize the performance of the material, and improve prediction accuracy and design efficiency.

[0051] From the above description, it can be seen that the above-mentioned embodiments of the present invention achieve the following technical effects: the method used to predict and design carbon dot-based long afterglow materials is highly efficient, can quickly screen out the best combination of ingredients, and reduce the number of experiments; can optimize the performance of the material, and achieve better luminescence performance by precisely adjusting the ingredients; can save costs, reduce the time and economic cost of material research and development; and can achieve accurate predictions, improve the accuracy and reliability of predictions, and reduce the trial and error process.

[0052] Those skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the present invention (including the claims) is limited to these examples. Under the concept of the present invention, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of the different aspects of the present invention as described above, which are not provided in detail for the sake of simplicity.

[0053] The present invention is intended to cover all such substitutions, modifications and variations that fall within the broad scope of the claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. Using machine learning to predict and design carbon dot-based long afterglow materials, characterized in that: The specific steps include: S1: Data collection and organization: S1a: Experimental data: Collect the experimental data of existing carbon dot materials; S1b: Literature data: integrate the property data of carbon dot-based materials reported in relevant literature; S2: Feature selection and processing: S2a: Feature selection: extract key features that affect the long afterglow characteristics from the collected data; S2b: Data preprocessing: standardize and preprocess the data, clean and organize the data, remove noise and incomplete information, so as to facilitate machine learning modeling and split the data into training and test sets; S3: Machine Learning Modeling: S3a: Model selection: Selecting a suitable machine learning algorithm; S3b: Model training: Use the data from the training set to train the model so that it can learn the relationship between the properties of carbon dot-based long afterglow materials and their composition; S3c: Model evaluation: Use the data from the test set to evaluate the predictive ability of the model and use appropriate evaluation metrics for validation; S4: Prediction and Optimization: S4a: Prediction: Use the trained model to predict the performance of new carbon dot-based long afterglow materials, determine which components and conditions may lead to excellent long afterglow effects, explore different combinations of material components and structures, and optimize the long afterglow characteristics; S4b: Optimization: Optimization algorithms are used to further adjust the composition and preparation conditions to design carbon dot materials with the best performance; S5: Verification and Experimentation: S5a: Model verification: Apply the results predicted by the model to actual experiments, synthesize and characterize the predicted carbon dot-based long afterglow materials, and verify the accuracy of its predictions; S5b: Experimental adjustment: Adjust the model and design scheme according to the experimental results, perform iterative optimization, gradually optimize the performance of the material, and improve prediction accuracy and design efficiency.

2. The method for predicting and designing carbon dot-based long afterglow materials using machine learning according to claim 1, characterized in that: In S1a, the experimental data include composition, structure, performance, preparation conditions and optical properties.

3. The method for predicting and designing carbon dot-based long afterglow materials using machine learning according to claim 2, characterized in that: Optical properties include luminous intensity and duration.

4. The method for predicting and designing carbon dot-based long afterglow materials using machine learning according to claim 1, characterized in that: In S2a, key features include the size of carbon dots, surface functional groups, doping elements, and synthesis method parameters.

5. The method for predicting and designing carbon dot-based long afterglow materials using machine learning according to claim 1, characterized in that: In S3a, the machine learning algorithm is such as a regression model, a classification model or a deep learning model.

6. The method for predicting and designing carbon dot-based long afterglow materials using machine learning according to claim 5, characterized in that: Regression models such as linear regression and support vector regression, classification models such as decision trees and random forests, and deep learning models such as neural networks.

7. The method of predicting and designing carbon dot-based long afterglow materials using machine learning according to claim 1, characterized in that: In S3c, the evaluation indicators include mean square error and accuracy.

8. The method of predicting and designing carbon dot-based long afterglow materials using machine learning according to claim 1, characterized in that: In S4b, the optimization algorithm is, for example, a genetic algorithm or a particle swarm optimization.

Images