Method and system for predicting stability of lead-free double perovskite halide

Through machine learning methods, screen the characteristic terms and Pearson correlation coefficients of double perovskite materials, build a model to predict their formation energy value, solve the problem of low stability judgment efficiency in the existing technology, and achieve high precision material performance prediction and rapid screening.

CN119069024APending Publication Date: 2024-12-03BEIJING UNIV OF POSTS & TELECOMM
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202410960829.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-17
Publication Date
2024-12-03

AI Technical Summary

Technical Problem

The prior art requires processing large-scale complex data when determining the stability of double perovskite materials, which is low in efficiency, making it difficult to quickly screen out new materials with ideal light absorption bandwidth and good stability.

Method used

Using machine learning methods, by screening feature items from the relevant data of lead-free double perovskite halides, calculating Pearson correlation coefficients, constructing a model to predict the formation energy value of the material, determining its stability, and using a database to identify difficult humans to discover the pattern and relationship to improve the accuracy of material performance prediction.

Benefits of technology

The cycle from new material design to verification is greatly reduced, providing high-precision material performance prediction, and improving the stability prediction efficiency and accuracy of lead-free bisovskite halides.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119069024A_ABST
    Figure CN119069024A_ABST
Patent Text Reader

Abstract

The invention provides a lead-free double perovskite halide stability prediction method and system. The method comprises the following steps: obtaining related data of a to-be-measured lead-free double perovskite halide from a data source; screening out data of a plurality of feature items from related data of the lead-free double perovskite halide, and combining every two feature items into a feature combination; calculating Pearson's correlation coefficients of the two feature items in each feature combination, and screening all the calculated Pearson's correlation coefficients; evaluating the performance of the plurality of models based on a test set corresponding to the to-be-measured lead-free double perovskite halide, and selecting the models; and constructing input data based on the screened Pearson's correlation coefficients, inputting the input data into a selected model, outputting a formation energy value of the to-be-measured lead-free double perovskite halide by the selected model, and judging whether the to-be-measured lead-free double perovskite halide is stable or not based on the formation energy value.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence and material technology, and in particular to a method and system for predicting the stability of lead-free double perovskite halides. Background Art

[0002] Double perovskite materials have excellent optical absorption ability, long carrier diffusion length and excellent properties that can be processed by solution, showing great potential in the field of optoelectronics. Compared with traditional single perovskite materials, double perovskite materials are widely used in photovoltaic devices, light-emitting diodes, photocatalysis, spin electronics, sensors, thermoelectric materials, nonlinear optical materials and energy storage materials due to their excellent photoelectric conversion efficiency, optical properties, photocatalytic properties, magnetic and electrical transport properties. Therefore, in-depth research on non-toxic and efficient double perovskite materials has important theoretical value and broad application prospects. Due to the flexible elemental composition and diverse crystal structure of double perovskite materials, there are many types that can be designed. Therefore, it has become the main focus of current research to quickly screen out a new material with ideal light absorption bandwidth and good stability from many double perovskite materials.

[0003] However, determining the stability of double perovskite materials often requires processing large amounts of complex data, which is inefficient. Summary of the invention

[0004] In view of this, an embodiment of the present invention provides a method for predicting the stability of lead-free double perovskite halide to eliminate or improve one or more defects existing in the prior art.

[0005] One aspect of the present invention provides a method for predicting the stability of lead-free double perovskite halides, the method comprising the following steps:

[0006] Obtaining relevant data of the lead-free double perovskite halide to be measured from a data source;

[0007] Screening out data of multiple characteristic items from relevant data of lead-free double perovskite halides, and combining every two characteristic items into a characteristic combination;

[0008] Calculate the Pearson correlation coefficient of the two feature items in each feature combination, and screen all the calculated Pearson correlation coefficients;

[0009] The performance of multiple models is evaluated based on a test set corresponding to the lead-free double perovskite halides to be measured, and a model is selected;

[0010] Input data is constructed based on the screened Pearson correlation coefficient and input into the selected model. The selected model outputs the formation energy value of the lead-free double perovskite halide to be measured, and whether the lead-free double perovskite halide to be measured is stable is determined based on the formation energy value.

[0011] The above scheme adopts the machine learning method to determine whether the lead-free double perovskite halide to be measured is stable. This combination of machine learning algorithm and database can identify patterns and relationships that are difficult for humans to discover, and can provide high-precision material performance predictions, greatly shortening the cycle from new material design to verification, which is for lead-free double perovskite halide.

[0012] In some embodiments of the present invention, the performance of multiple models is evaluated based on a test set corresponding to the lead-free double perovskite halide to be measured, and the step of selecting a model includes: selecting multiple models based on the mean square error, root mean square error and determination coefficient of the model, and selecting a model.

[0013] In some embodiments of the present invention, in the step of screening out data of multiple feature items from relevant data of lead-free double perovskite halides, the SHAP value of each feature item is calculated, and a preset number of feature items are screened out based on the SHAP value.

[0014] In some embodiments of the present invention, in the step of calculating the Pearson correlation coefficient of the two feature items in each feature combination, the Pearson correlation coefficient of the two feature items is calculated based on the following formula:

[0015]

[0016] Among them, X i and Y i Represent the observed values ​​of the two feature items respectively, and They represent the means of the two characteristic items, n is the sample size, and r represents the Pearson correlation coefficient.

[0017] In some embodiments of the present invention, in the step of screening all calculated Pearson correlation coefficients, the Pearson correlation coefficients are screened based on a preset screening threshold, and the Pearson correlation coefficients greater than the screening threshold are deleted.

[0018] In some embodiments of the present invention, in the step of determining whether the lead-free double perovskite halide to be measured is stable based on the formation energy value, a preset determination threshold is compared with the formation energy value to determine whether the lead-free double perovskite halide to be measured is stable.

[0019] In some embodiments of the present invention, in the step of constructing input data based on the screened Pearson correlation coefficient, the importance value of each feature combination is constructed based on the SHAP value of the feature item, the Pearson correlation coefficient and the importance value of each feature combination are taken as a numerical group, and the numerical group of each feature combination is constructed as input data.

[0020] In some embodiments of the present invention, in the step of constructing the importance value of each feature combination based on the SHAP values ​​of the feature items, the average value of the SHAP values ​​of the two feature items in the feature combination is calculated as the importance value of the feature combination.

[0021] In some embodiments of the present invention, in the step of selecting a model from multiple models based on the mean square error, root mean square error and determination coefficient of the models, the weighted average of the three items of data, mean square error, root mean square error and determination coefficient, is calculated, and the model is selected based on the calculated weighted average.

[0022] The second aspect of the present invention also provides a lead-free double perovskite halide stability prediction system, which includes a computer device, the computer device includes a processor and a memory, the memory stores computer instructions, the processor is used to execute the computer instructions stored in the memory, and when the computer instructions are executed by the processor, the system implements the steps implemented by the method described above.

[0023] The third aspect of the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps implemented by the aforementioned lead-free double perovskite halide stability prediction method.

[0024] Additional advantages, purposes, and features of the present invention will be described in part in the following description, and will become apparent to those skilled in the art after studying the following, or may be learned from the practice of the present invention. The purposes and other advantages of the present invention can be specifically pointed out and obtained in the specification and the accompanying drawings.

[0025] Those skilled in the art will appreciate that the objectives and advantages that can be achieved with the present invention are not limited to the above specific description, and the above and other objectives that can be achieved by the present invention will be more clearly understood from the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The drawings described herein are used to provide a further understanding of the present invention, constitute a part of this application, and do not constitute a limitation of the present invention.

[0027] Figure 1 A schematic diagram of an implementation of a method for predicting stability of lead-free double perovskite halides according to the present invention;

[0028] Figure 2 A schematic diagram of the overall architecture of the present invention;

[0029] Figure 3 Schematic diagram of the calculated Pearson correlation coefficient. DETAILED DESCRIPTION

[0030] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments and the accompanying drawings. Here, the illustrative embodiments of the present invention and their descriptions are used to explain the present invention, but are not intended to limit the present invention.

[0031] It should also be noted that, in order to avoid obscuring the present invention due to unnecessary details, only structures and / or processing steps closely related to the solutions according to the present invention are shown in the accompanying drawings, while other details that are not closely related to the present invention are omitted.

[0032] Specifically, lead-free double perovskite halides are a new type of material formed by replacing the Pb element in traditional lead-based perovskites with non-toxic candidate elements (such as Cs, Ag, Bi, Br, etc.). Its chemical formula is usually A 2 B'B"X 6 (where A is a monovalent cation, such as Cs; B' and B" are divalent and trivalent metal cations, such as Ag and Bi; and X is a halogen anion, such as Br). This structure is similar to traditional perovskites, but has higher complexity and diversity.

[0033] Lead-free double perovskite halides usually have low thermal conductivity, which is of great significance for thermoelectric conversion and thermal management of electronic devices. For example, the team of Professor Huang Baoling of the Hong Kong University of Science and Technology used the Tianhe-2 supercomputer to study the lead-free halide double perovskite Cs 2 AG 6 , and found that it has ultra-low thermal conductivity. This discovery provides a scientific basis for the development of thermoelectric materials and thermal management of solar cells; lead-free double perovskite halides also show excellent performance in the optoelectronic field. They can efficiently absorb and emit photons and are suitable for optoelectronic devices such as solar cells, light-emitting diodes (LEDs), and photodetectors. In particular, their environmental friendliness and stability make them an ideal choice to replace traditional lead-based perovskite materials; lead-free double perovskite halides generally have good chemical stability and thermal stability, and can maintain the stability of their structure and performance over a wide temperature range.

[0034] Lead-free double perovskite halides as absorber materials for solar cells have high photoelectric conversion efficiency and good stability. 2 AG 6Based on perovskite solar cells, high power conversion efficiency has been achieved, and good stability is achieved without encapsulation under ambient conditions; lead-free double perovskite halides can also be used to manufacture light-emitting diodes. By doping different elements or ions, different colors of light can be achieved to meet different application requirements; due to its excellent optoelectronic properties and stability, lead-free double perovskite halides are also used to manufacture optoelectronic devices such as photodetectors.

[0035] like Figure 1 and 2 As shown, the present invention proposes a method for predicting the stability of lead-free double perovskite halides, which comprises the following steps:

[0036] Step S100, obtaining relevant data of the lead-free double perovskite halide to be measured from a data source;

[0037] In the specific implementation process, the step of obtaining relevant data of the lead-free double perovskite halide to be measured from the data source can obtain the required data from the Materials Project through a python program. The parameters of the required data can be modified in the program to stably capture the data in the Materials Project, and then the Pandas library is used for data processing to extract characteristic items such as ionic radius, tolerance factor and octahedral factor.

[0038] Using the above scheme, relevant data of lead-free double perovskite halides are collected from public databases or literature, including but not limited to chemical composition, ionic radius, formation energy, tolerance factor and octahedral factor. The data are organized into a structured table format and missing values ​​are processed. Feature extraction and selection are performed on the collected data to identify the features that have the greatest impact on the prediction of material stability.

[0039] Step S200, filtering out data of a plurality of characteristic items from relevant data of lead-free double perovskite halides, and combining every two characteristic items into a characteristic combination;

[0040] In some embodiments of the present invention, in the step of screening out data of multiple characteristic items from relevant data of lead-free double perovskite halides, the screened out multiple characteristic items may be density, atomic density, atomic energy, band gap, Fermi level, ionic radius of A-site element, ionic radius of B'-site element, ionic radius of B"-site element, tolerance factor, new tolerance factor, octahedral factor, atomic number of B'-site element, mass number of B'-site element, electron affinity of B'-site element, van der Waals radius of B'-site element, boiling point of B'-site element, atomic number of B"-site element, mass number of B"-site element, electronegativity of B"-site element, electron affinity of B"-site element, van der Waals radius of B"-site element, boiling point of B"-site element, ionization energy of X-site element and electron affinity of X-site element.

[0041] like Figure 3 As shown, step S300, calculate the Pearson correlation coefficients of the two feature items in each feature combination, and screen all the calculated Pearson correlation coefficients;

[0042] In the specific implementation process, the Pearson correlation coefficient (Pearson correlation coefficient), also known as the Pearson product-moment correlation coefficient (PPMCC or PCCs for short), is a statistical indicator used to measure the degree of linear correlation between two variables, and its value is between -1 and 1.

[0043] Step S400, evaluating the performance of multiple models based on a test set corresponding to the lead-free double perovskite halide to be measured, and selecting a model;

[0044] In the specific implementation process, the performance of multiple models is evaluated based on the test set corresponding to the lead-free double perovskite halide to be measured. In the step of selecting the model, the multiple models are all pre-trained XGBoost algorithm models. Specifically, the multiple models are trained using different training data sets.

[0045] In the specific implementation process, the model can be selected based on the accuracy of the test set.

[0046] In the specific implementation process, the core principle of the XGBoost algorithm is to build a tree model by optimizing the second-order expansion of the loss function. XGBoost controls the complexity of the model by introducing regularization and pruning techniques, thereby improving the generalization ability and prediction accuracy of the model. It can handle large-scale data and high-dimensional features, and supports parallel computing and distributed training, making it perform well in a big data environment.

[0047] In a specific implementation process, the test set corresponding to the lead-free double perovskite halide to be measured may be data obtained by screening all calculated Pearson correlation coefficients, or may be a pre-set data set.

[0048] Step S500, constructing input data based on the screened Pearson correlation coefficient and inputting it into the selected model, the selected model outputs the formation energy value of the lead-free double perovskite halide to be measured, and judging whether the lead-free double perovskite halide to be measured is stable based on the formation energy value.

[0049] In some embodiments of the present invention, the formation energy of the lead-free double perovskite halide is a complex physicochemical parameter, which involves multiple aspects such as the stability of the compound, the crystal structure, and the interaction between elements. However, the specific value of the formation energy usually needs to be determined by complex theoretical calculations (such as density functional theory DFT) or experimental measurements.

[0050] The above scheme uses machine learning methods to predict the formation energy, and then determines whether the lead-free double perovskite halide to be measured is stable. This combination of machine learning algorithms and databases can identify patterns and relationships that are difficult for humans to discover, and can provide high-precision material performance predictions, greatly shortening the cycle from new material design to verification, which is for lead-free double perovskite halides.

[0051] In some embodiments of the present invention, the performance of multiple models is evaluated based on a test set corresponding to the lead-free double perovskite halide to be measured, and the step of selecting a model includes: selecting multiple models based on the mean square error, root mean square error and determination coefficient of the model, and selecting a model.

[0052] In some embodiments of the present invention, in the step of selecting a model from multiple models based on the mean square error, root mean square error and determination coefficient of the models, a single item comparison can be performed on multiple models based on the selection criteria of any one of the items; or the mean square error, root mean square error and determination coefficient can be combined to select a model from multiple items.

[0053] In some embodiments of the present invention, in the step of screening out data of multiple feature items from relevant data of lead-free double perovskite halides, the SHAP value of each feature item is calculated, and a preset number of feature items are screened out based on the SHAP value.

[0054] In its implementation, SHAP values ​​generate a value (also called a SHAP value) for each input feature that indicates how that feature contributes to the prediction of a given data point. Some features contribute positively to the prediction, while others contribute negatively.

[0055] In some embodiments of the present invention, in the step of calculating the Pearson correlation coefficient of the two feature items in each feature combination, the Pearson correlation coefficient of the two feature items is calculated based on the following formula:

[0056]

[0057] Among them, X i and Y i Represent the observed values ​​of the two feature items respectively, and They represent the means of the two characteristic items, n is the sample size, and r represents the Pearson correlation coefficient.

[0058] In some embodiments of the present invention, in the step of screening all calculated Pearson correlation coefficients, the Pearson correlation coefficients are screened based on a preset screening threshold, and the Pearson correlation coefficients greater than the screening threshold are deleted.

[0059] In a specific implementation process, in the step of screening the Pearson correlation coefficient based on a preset screening threshold and deleting the Pearson correlation coefficient greater than the screening threshold, the screening threshold may be 0.7, 0.8 or 0.9, preferably 0.8.

[0060] In some embodiments of the present invention, in the step of determining whether the lead-free double perovskite halide to be measured is stable based on the formation energy value, a preset determination threshold is compared with the formation energy value to determine whether the lead-free double perovskite halide to be measured is stable.

[0061] The above scheme is adopted. This scheme is based on comparing the predicted formation energy with the judgment threshold. Since the formation energy involves multiple aspects such as the stability of the compound, the crystal structure, and the interaction between elements, the formation energy is used to determine whether the lead-free double perovskite halide to be measured is stable.

[0062] In some embodiments of the present invention, in the step of constructing input data based on the screened Pearson correlation coefficient, the importance value of each feature combination is constructed based on the SHAP value of the feature item, the Pearson correlation coefficient and the importance value of each feature combination are taken as a numerical group, and the numerical group of each feature combination is constructed as input data.

[0063] In some embodiments of the present invention, in the step of constructing the importance value of each feature combination based on the SHAP values ​​of the feature items, the average value of the SHAP values ​​of the two feature items in the feature combination is calculated as the importance value of the feature combination.

[0064] Adopting the above scheme, this scheme combines the Pearson correlation coefficient and importance value of each feature combination as a numerical group, wherein the importance value combines the SHAP values ​​of the two feature items. Since the SHAP value can indicate how the feature contributes to the prediction of a specified data point, and the Pearson correlation coefficient can measure the degree of linear correlation between two variables, this scheme combines the Pearson correlation coefficient and importance value of each feature combination, and can predict the formation energy value by measuring the importance and correlation of the two feature items in the combination. Compared with the method of using the Pearson correlation coefficient alone, it can better ensure the accuracy of the prediction.

[0065] In some embodiments of the present invention, in the step of selecting a model from multiple models based on the mean square error, root mean square error and determination coefficient of the models, the weighted average of the three items of data, mean square error, root mean square error and determination coefficient, is calculated, and the model is selected based on the calculated weighted average.

[0066] An embodiment of the present invention also provides a lead-free double perovskite halide stability prediction system, which includes a computer device, wherein the computer device includes a processor and a memory, wherein the memory stores computer instructions, and the processor is used to execute the computer instructions stored in the memory. When the computer instructions are executed by the processor, the system implements the steps implemented by the method described above.

[0067] The embodiment of the present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps implemented by the aforementioned lead-free double perovskite halide stability prediction method are implemented. The computer-readable storage medium can be a tangible storage medium, such as a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a floppy disk, a hard disk, a removable storage disk, a CD-ROM, or any other form of storage medium known in the technical field.

[0068] It should be understood by those skilled in the art that the exemplary components, systems and methods described in conjunction with the embodiments disclosed herein can be implemented in hardware, software or a combination of the two. Whether it is performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present invention are programs or code segments used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link via a data signal carried in a carrier.

[0069] It should be clear that the present invention is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present invention.

[0070] In the present invention, features described and / or illustrated for one embodiment may be used in the same or similar manner in one or more other embodiments, and / or combined with features of other embodiments or replace features of other embodiments.

[0071] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the embodiments of the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for predicting the stability of lead-free double perovskite halides, characterized in that: The method comprises the following steps: Obtaining relevant data of the lead-free double perovskite halide to be measured from a data source; Screening out data of multiple characteristic items from relevant data of lead-free double perovskite halides, and combining every two characteristic items into a characteristic combination; Calculate the Pearson correlation coefficient of the two feature items in each feature combination, and screen all the calculated Pearson correlation coefficients; The performance of multiple models is evaluated based on a test set corresponding to the lead-free double perovskite halides to be measured, and a model is selected; Input data is constructed based on the screened Pearson correlation coefficient and input into the selected model. The selected model outputs the formation energy value of the lead-free double perovskite halide to be measured, and whether the lead-free double perovskite halide to be measured is stable is determined based on the formation energy value.

2. The method for predicting the stability of lead-free double perovskite halides according to claim 1, characterized in that: The performance of multiple models is evaluated based on the test set corresponding to the lead-free double perovskite halide to be measured, and the step of selecting a model includes: selecting multiple models based on the mean square error, root mean square error and determination coefficient of the model, and selecting a model.

3. The method for predicting the stability of lead-free double perovskite halides according to claim 1, characterized in that: In the step of screening out data of a plurality of characteristic items from relevant data of lead-free double perovskite halides, a SHAP value of each characteristic item is calculated, and a preset number of characteristic items are screened out based on the SHAP value.

4. The method for predicting the stability of lead-free double perovskite halides according to claim 1, characterized in that: In the step of calculating the Pearson correlation coefficient of the two feature items in each feature combination, the Pearson correlation coefficient of the two feature items is calculated based on the following formula: Among them, X i and Y i Represent the observed values ​​of the two feature items respectively, and They represent the means of the two characteristic items, n is the sample size, and r represents the Pearson correlation coefficient.

5. The method for predicting the stability of lead-free double perovskite halides according to claim 3, characterized in that: In the step of screening all calculated Pearson correlation coefficients, the Pearson correlation coefficients are screened based on a preset screening threshold, and the Pearson correlation coefficients greater than the screening threshold are deleted.

6. The method for predicting the stability of lead-free double perovskite halides according to claim 3, characterized in that: In the step of determining whether the lead-free double perovskite halide to be measured is stable based on the formation energy value, a preset determination threshold is compared with the formation energy value to determine whether the lead-free double perovskite halide to be measured is stable.

7. The method for predicting the stability of lead-free double perovskite halides according to claim 1, characterized in that: In the step of constructing input data based on the filtered Pearson correlation coefficient, the importance value of each feature combination is constructed based on the SHAP value of the feature item, the Pearson correlation coefficient and the importance value of each feature combination are taken as a value group, and the value group of each feature combination is constructed as input data.

8. The method for predicting the stability of lead-free double perovskite halides according to any one of claims 1 to 7, characterized in that: In the step of constructing the importance value of each feature combination based on the SHAP values ​​of the feature items, an average value of the SHAP values ​​of the two feature items in the feature combination is calculated as the importance value of the feature combination.

9. The method for predicting the stability of lead-free double perovskite halides according to claim 8, characterized in that: In the step of selecting a model from multiple models based on the mean square error, root mean square error and determination coefficient of the models, a weighted average of the mean square error, root mean square error and determination coefficient is calculated, and the model is selected based on the calculated weighted average.

10. A lead-free double perovskite halide stability prediction system, characterized in that: The system includes a computer device, which includes a processor and a memory. The memory stores computer instructions. The processor is used to execute the computer instructions stored in the memory. When the computer instructions are executed by the processor, the system implements the steps implemented by the method as described in any one of claims 1 to 9.

Citation Information

Patent Citations

  • Method for efficiently predicting stability of perovskite based on integrated machine learning

    CN113052367A