CN114974476A - Method for selecting perovskite battery modification material - Google Patents
Method for selecting perovskite battery modification material Download PDFInfo
- Publication number
- CN114974476A CN114974476A CN202210569200.3A CN202210569200A CN114974476A CN 114974476 A CN114974476 A CN 114974476A CN 202210569200 A CN202210569200 A CN 202210569200A CN 114974476 A CN114974476 A CN 114974476A
- Authority
- CN
- China
- Prior art keywords
- machine learning
- learning algorithm
- perovskite
- data set
- characteristic
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
- G16C60/00—Computational materials science, i.e. ICT specially adapted for investigating the physical or chemical properties of materials or phenomena associated with their design, synthesis, processing, characterisation or utilisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/285—Selection of pattern recognition techniques, e.g. of classifiers in a multi-classifier system
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E10/00—Energy generation through renewable energy sources
- Y02E10/50—Photovoltaic [PV] energy
- Y02E10/549—Organic PV cells
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Biology (AREA)
- Software Systems (AREA)
- Medical Informatics (AREA)
- Mathematical Physics (AREA)
- Photovoltaic Devices (AREA)
Abstract
The invention provides a selection method of a perovskite battery modification material. The method comprises the following steps: obtaining modification material information, perovskite precursor characteristics and device performance parameters before and after optimization of the modification material adopted by the existing perovskite battery, and establishing an initial data characteristic set; carrying out dimensionality reduction on the initial data feature set to obtain an important feature data set; training and testing different machine learning algorithm models by using the important characteristic data set, and selecting an optimized machine learning algorithm model; screening key features in the important feature data set by using the optimization model, and establishing a mapping model between the key features and the final device performance of the perovskite battery; and selecting a corresponding high-efficiency modified material according to the performance of the high-efficiency perovskite battery device predicted by the mapping model. The invention can quantify the influence of each characteristic of the modified material on the output performance of the perovskite battery, and effectively selects the high-efficiency modified material according to the performance of the high-efficiency perovskite battery device required to be realized.
Description
Technical Field
The invention relates to the technical field of perovskite battery material data mining, in particular to a method for selecting a perovskite battery modification material.
Background
Metal halide Perovskite Solar Cells (PSCs) have shown impressive Power Conversion Efficiency (PCE) in the past decade, making them one of the most promising photovoltaic technologies. The single-junction PSC has reached a certified PCE of 25.7% over other thin-film solar cells. The high PCE results from the excellent optoelectronic properties of the perovskite. However, the presence of defects makes the PCE of perovskite solar cells still well below the Shockley-Queisser limit. The presence of many point defects at the grain boundaries and interfaces of perovskite thin films traps photogenerated carriers and degrades device performance (e.g., uncoordinated Pb 2+ And moving I - ). These defects originate mainly from the polycrystalline and soft structure of the perovskite. To passivate these defects, researchers have proposed two general approaches. One is to introduce a suitable interface material on top of the perovskite layer and the other is to add suitable additives to the perovskite precursor solution. Both passivation strategies have proven to be effective in increasing the PCE of perovskite solar cells.
Over the past few years, a large number of interface materials and additives have been reported to modify perovskites. Several mechanisms are proposed to explain their function, including their functional groups with uncoordinated Pb 2+ The formation of wide band gap two-dimensional perovskites, etc. These mechanisms help researchers make qualitative judgments about the function of interface materials; however, there is still a lack of general guidelines for intelligent screening or design of materials. For example, a researcher may desire to have NH 3 + The materials at the tip (e.g., PEAI) improve device performance, but they cannot determine if it is more effective than PEAI. At the same time, it is difficult to compare the work of different groups of functional groupsAnd (5) effect. Therefore, in the search for high performance interface materials and additives, it would be of great interest if the relationship between the material/additive chemical structure and the photovoltaic characteristics of the perovskite cell could be established prior to device fabrication.
Metal halide Perovskite Solar Cells (PSCs) have shown impressive Power Conversion Efficiency (PCE) in the past decade, making them one of the most promising photovoltaic technologies. The single-junction PSC has reached a certified PCE of 25.7% over other thin-film solar cells. The high PCE results from the excellent optoelectronic properties of the perovskite. However, the presence of defects makes the PCE of perovskite solar cells still well below the Shockley-Queisser limit. The presence of many point defects at the grain boundaries and interfaces of perovskite thin films traps photogenerated carriers and degrades device performance (e.g., uncoordinated Pb 2+ And moving I - ). These defects originate mainly from the polycrystalline and soft structure of the perovskite. To passivate these defects, researchers have proposed two general approaches. One is to introduce a suitable interface material on top of the perovskite layer and the other is to add suitable additives to the perovskite precursor solution. Both passivation strategies have proven effective in increasing the PCE of perovskite solar cells.
During the last few years, a large number of interface materials and additives have been reported to modify perovskites. Several mechanisms are proposed to explain their function, including their functional groups with uncoordinated Pb 2+ The formation of wide band gap two-dimensional perovskites, etc. These mechanisms help researchers make qualitative judgments about the function of interface materials; however, there is still a lack of general guidelines for intelligent screening or design of materials. For example, a researcher may desire to have NH 3 + The materials at the tip (e.g., PEAI) improve device performance, but they cannot determine if it is more effective than PEAI. At the same time, it is difficult to compare the efficacy of different groups of functional groups. Therefore, in the search for high performance interface materials and additives, it would be of great interest if the relationship between the material/additive chemical structure and the photovoltaic characteristics of the perovskite cell could be established prior to device fabrication.
Disclosure of Invention
The embodiment of the invention provides a method for selecting a perovskite battery modified material, so that the efficient modified material can be effectively selected according to the performance of a high-efficiency perovskite battery device required to be realized.
In order to achieve the purpose, the invention adopts the following technical scheme.
A method of selecting a perovskite battery modification material, comprising:
acquiring modification material information, perovskite precursor characteristics and device performance parameters before and after optimization of the modification material adopted by the existing perovskite battery, taking the modification material characteristics, the perovskite precursor characteristics and the device performance parameters before optimization of the modification material as input independent variables, and taking the device photovoltaic characteristic parameters after optimization of the modification material as output dependent variables, and establishing an initial data characteristic set;
performing dimensionality reduction on the initial data feature set to obtain an important feature data set;
establishing different machine learning algorithm models, training and testing the different machine learning algorithm models by using the important characteristic data set, evaluating the performance of each machine learning algorithm model, and selecting an optimized machine learning algorithm model;
analyzing and screening key features of the modified materials in the important feature data set by using an optimized machine learning algorithm model, combining the key features to construct the modified materials, predicting performance parameters of the constructed modified materials by using the optimized machine learning algorithm model, and establishing a mapping model between the key features and the final device performance of the perovskite battery by using a machine learning method;
and selecting a modification material according to the perovskite battery device performance required to be realized by utilizing the mapping model.
Preferably, the perovskite cell comprises a perovskite solar cell.
Preferably, the method comprises the steps of obtaining the information of the modified material adopted by the existing perovskite battery, the characteristics of the perovskite precursor and the performance parameters of the device before and after optimization of the modified material, taking the characteristics of the modified material, the characteristics of the perovskite precursor and the performance parameters of the device before optimization of the modified material as input independent variables, and taking the photovoltaic characteristic parameters of the device after optimization of the modified material as output dependent variables, and establishing an initial data feature set comprising;
extracting a modification material adopted by the perovskite solar cell in a literature report/experimental report, and obtaining SMILES codes of the modification material by adopting molecular editing software or a webpage;
based on SMILES coding, extracting the characteristics of the modified material, then extracting perovskite precursor characteristics and device performance parameters before and after optimization of the modified material, taking the modified material characteristics, the perovskite precursor characteristics and the device performance parameters before optimization of the modified material as input independent variables, and taking the device photovoltaic characteristic parameters after optimization of the modified material as output dependent variables, and establishing an initial data feature set.
Preferably, the dimension reduction processing is performed on the initial data feature set to obtain an important feature data set, including;
preprocessing the initial set of data features, the preprocessing comprising: cleaning error data and abnormal value data in the initial data feature set, and supplementing some feature missing values in the initial data feature set to obtain a cleaned data set;
and (3) reducing the dimensions of the cleaned data set by adopting correlation analysis, a principal component analysis method, a self-coding grid method or discrete point analysis to obtain an important characteristic data set.
Preferably, the establishing of different machine learning algorithm models, the training and testing of different machine learning algorithm models by using the important feature data set, the evaluation of the performance of each machine learning algorithm model, and the selection of an optimized machine learning algorithm model comprises;
establishing different machine learning algorithm models by adopting linear regression, random forests, extreme gradient boosting and neural network algorithms;
dividing the important feature data set into a training set and a testing set, training and testing different machine learning algorithm models by using the training set and the testing set, evaluating the performance of each machine learning algorithm model by using a Root Mean Square Error (RMSE) and a Pearson correlation coefficient r, screening the machine learning algorithm model which executes the lowest RMSE and the highest r value on the testing set, and performing subsequent prediction by using the machine learning algorithm model as an optimized machine learning algorithm model;
the index formula for calculating RMSE is as follows:
the formula for calculating the Pearson correlation coefficient r is as follows:
wherein X i ,Y i ,
And n represents the ith value of the experimental data set, the ith value of the predicted data set, the average of the experimental data set, the average of the predicted data set, and the number of experimental data set points, respectively.
Preferably, the key features of the modified material in the important feature data set are analyzed and screened by using the optimized machine learning algorithm model, the key features are combined to construct the modified material, and the optimized machine learning algorithm model is adopted to predict the performance parameters of the constructed modified material, including;
analyzing and quantifying the influence of each characteristic in the important characteristic data set on output by using the importance analysis of the screened optimized machine learning algorithm model, and screening the most important key characteristic of the modified material;
combining the screened key characteristics to construct a modified material of the perovskite battery, obtaining a SMILES code of the constructed modified material by using molecular editing software or a webpage, extracting the characteristics of the constructed modified material based on the SMILES code, and then establishing a prediction data characteristic set based on different perovskite precursor characteristics and device performance parameters before optimization of different modified materials;
and predicting the prediction data characteristic set by adopting an optimized machine learning algorithm to obtain corresponding output values of the modified material of the perovskite battery under different perovskite precursor characteristics and device performance parameters before optimization of different modified materials, wherein the output values comprise device performance parameter values or perovskite characteristic parameter values, and establishing a mapping model between the key characteristics and the final device performance of the perovskite battery by utilizing a machine learning method.
Preferably, the method further comprises;
corresponding experimental verification is carried out on the perovskite battery modified material selected according to the prediction result of the optimized machine learning algorithm model, corresponding theoretical verification is carried out by adopting device simulation, first principle calculation and related theories, and the accuracy of the optimized machine learning algorithm model is evaluated.
It can be seen from the technical solutions provided by the embodiments of the present invention that the method according to the embodiments of the present invention enables researchers to quantify the influence of each characteristic of the modified material on the output performance of the perovskite cell in advance in the process of searching for high-performance interface materials and additives, screens out the most important characteristic of the modified material, and can establish the relationship between the chemical structure of the material/additive and the photovoltaic characteristics of the perovskite cell before the device is prepared.
Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
FIG. 1 is a process flow diagram of a method for selecting a perovskite battery modification material according to an embodiment of the invention;
FIG. 2 is a graph illustrating an importance analysis of input feature values according to an embodiment of the present invention;
fig. 3 is a comparison between predicted values and actual values in a training set and a testing set based on different algorithm models according to an embodiment of the present invention.
Detailed Description
Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the accompanying drawings are illustrative only for the purpose of explaining the present invention, and are not to be construed as limiting the present invention.
As used herein, the singular forms "a", "an", "the" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and/or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof. It will be understood that when an element is referred to as being "connected" or "coupled" to another element, it can be directly connected or coupled to the other element or intervening elements may also be present. Further, "connected" or "coupled" as used herein may include wirelessly connected or coupled. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items.
It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
For the convenience of understanding the embodiments of the present invention, the following description will be further explained by taking several specific embodiments as examples in conjunction with the drawings, and the embodiments are not to be construed as limiting the embodiments of the present invention.
The embodiment provides a method for selecting a perovskite cell modification material, which enables researchers to quantify the influence of each characteristic of the modification material on the output performance of a perovskite cell in advance in the process of searching for a high-performance interface material and an additive, screens out the most important characteristic of the modification material, and can establish the relationship between the chemical structure of the material/additive and the photovoltaic characteristic of the perovskite cell before device preparation.
The processing flow of the method for selecting the perovskite battery modification material provided by the embodiment of the invention is shown in fig. 1, and comprises the following processing steps:
and step S10, establishing an initial data feature set.
Obtaining modification material information, perovskite precursor characteristics and device performance parameters before and after optimization of the modification material adopted by the perovskite battery, which are recorded in a plurality of literature reports and experimental reports, taking the modification material characteristics, the perovskite precursor characteristics and the device performance parameters before optimization of the modification material as input independent variables, and taking the device photovoltaic characteristic parameters after optimization of the modification material as output dependent variables, and establishing an initial data characteristic set X. The perovskite cell may be a perovskite solar cell.
And S20, performing dimensionality reduction on the initial data feature set X by adopting correlation analysis, a principal component analysis method, a self-coding grid method and discrete point analysis to obtain an important feature data set X2.
And S30, establishing different machine learning algorithm models by adopting algorithms such as linear regression, random forest, extreme gradient promotion, neural network and the like, training and testing the different machine learning algorithm models by utilizing the important characteristic data set, evaluating the performance of each machine learning algorithm model by utilizing standard variance RMSE and Pearson correlation coefficient r, and selecting the optimized machine learning algorithm model.
And S40, performing weight analysis and screening on key features of the modified materials in the important feature data set by using the importance analysis function of the optimized machine learning algorithm model, combining the screened key features to construct the modified materials, predicting the constructed modified materials by using the optimized machine learning algorithm model, and establishing a mapping model between the key features and the final device performance of the perovskite battery by using a machine learning method.
And then, selecting a corresponding efficient modified material according to the performance of the high-efficiency perovskite battery device predicted by machine learning.
And step S50, developing corresponding experiments and theoretical calculation according to the model prediction result to verify the accuracy and the scientificity of the machine learning model analysis.
Specifically, step S10 includes:
s11, extracting the modified material adopted by the perovskite solar cell in the literature report/experimental report, and obtaining SMILES codes of the modified material by adopting molecular editing software or a webpage;
in this step, the SMILES (Simplified molecular input line entry System) code is a specification that explicitly describes the structure of a molecule using ACSII strings and is derived from the ChemDraw-20.0 software.
S12, extracting the characteristics of the modified material based on SMILES coding, wherein the characteristics include but are not limited to chemical informatics, electronic topological characteristics E-state-index and molecular fingerprint characteristics molecular fingerprints of the material, and then extracting perovskite precursor characteristics and device performance parameters before and after optimization of the modified material to establish an initial data characteristic set; wherein, the continuous variable adopts the value of the variable itself, and the discrete variable is represented by 0, 1, etc.
In this step, Chemnformatics was obtained from CDKDescUI-1.4.8 (java).
In this step, the E-state-index is obtained from PaDEL-Descriptor (java).
Specifically, step S20 includes:
s21, preprocessing: a large amount of obvious error data exists in the initial data feature set X, such as data value missing, abnormally large values and abnormally small values of a certain feature, the missing values of the feature are supplemented by common values, or discrete data values with the abnormally large values and the abnormally small values are cleaned in preprocessing, and finally a cleaned data set X1 is obtained;
s22, feature dimension reduction: and (3) performing dimensionality reduction on the cleaned data set X1 by adopting a correlation analysis method, a principal component analysis method, a self-coding grid method and a discrete point analysis method to obtain an important characteristic data set X2.
Specifically, step S30 includes:
s31, adopting Python or R language as a machine learning platform, and respectively establishing machine learning algorithm models such as linear regression, random forest, extreme gradient lifting, neural network and the like;
and S32, dividing the important characteristic data set X2 into a training set and a test set, adopting different machine learning algorithm models, and utilizing a ten-fold cross validation optimization algorithm hyper-parameter to ensure that the models present optimized performance in both the training set and the test set. And evaluating the performance of the model by using RMSE (Root Mean Square Error), Pearson correlation coefficient r and the like, screening a machine learning algorithm model executing the lowest RMSE and the highest r value on the test set, and performing subsequent prediction by using the machine learning algorithm model as an optimized machine learning algorithm model.
The index formula for calculating RMSE is as follows:
the formula for calculating the Pearson correlation coefficient r is as follows:
wherein X i ,Y i ,
And n represents the ith value of the experimental dataset, the ith value of the predicted dataset, the average of the experimental dataset, the average of the predicted dataset, and the number of experimental dataset points, respectively.
Fig. 2 is a comparison between predicted values and actual values in a training set and a testing set based on different machine learning algorithm models according to an embodiment of the present invention.
Specifically, step S40 includes:
s41, utilizing the importance analysis of the screened optimized machine learning algorithm model, carrying out weight analysis to quantify the influence of each characteristic in the important characteristic data set X2 on the output, and screening several most important key characteristics of the modified material. Fig. 3 is an importance analysis diagram of an input feature value according to an embodiment of the present invention.
S42, combining the screened key features to construct a modified material of the perovskite battery, obtaining a SMILES code of the constructed modified material by using ChemDraw-20.0 software, extracting features of the constructed modified material based on SMILES coding, wherein the features include but are not limited to chemical informatics, electronic topological features E-state-index and molecular fingerprint features molecular fingerprints of the constructed modified material, and then establishing a predicted data X feature set 3 based on different perovskite precursor features and device performance parameters before optimization of different modified materials;
s43, predicting the prediction data feature set X3 by adopting the screened optimized machine learning algorithm, and obtaining corresponding output values of the modified material of the perovskite battery under different perovskite precursor features and device performance parameters before optimization of different modified materials, wherein the output values comprise device performance parameter values or perovskite feature parameter values. And establishing a mapping model between the key characteristics and the final device performance of the perovskite battery by using a machine learning method. The mapping model includes the stoichiometric ratio in the perovskite precursor, and the functional group (NH) in the modified material 3 + S, -COOH, etc.), the length of the C chain, the complexity of the material, the hydrogen bond donor, etc., and the mapping between the final device performance of the perovskite cell.
And then, selecting a corresponding efficient modified material according to the performance of the high-efficiency perovskite battery device predicted by machine learning.
Specifically, step S50 includes:
s51, selecting high-efficiency perovskite battery modified materials according to the prediction result of the algorithm model, carrying out corresponding experimental verification, and evaluating the accuracy of the model;
and S52, analyzing the influence and effect of the screened modified material on the perovskite battery by adopting device simulation, first principle calculation and related theories, and proving the scientificity and rationality of a machine learning model and key characteristics.
In summary, the method for selecting a perovskite battery modification material provided in the embodiment of the present invention characterizes the modification material used in the preparation of the perovskite solar battery from a molecular perspective, and simultaneously characterizes the perovskite precursor and the device performance before and after the optimization of the modification material, and establishes a model between the perovskite battery device performance after the optimization of the modification material and the modification material characteristics, the perovskite precursor characteristics, and the device performance before the optimization of the modification material; furthermore, a mapping model between the characteristics and the final device performance of the perovskite battery is established by using a machine learning method, and the influence mechanism of the key characteristics on the perovskite thin film is clarified by combining a machine learning result and theoretical analysis, so that researchers can quantify the influence of each characteristic of the modified material on the output performance of the perovskite battery in advance in the process of searching for high-performance interface materials and additives, screen the most important characteristics of the modified material, and establish the relationship between the chemical structure of the material/additive and the photovoltaic characteristics of the perovskite battery before the device is prepared.
Those of ordinary skill in the art will understand that: the figures are merely schematic representations of one embodiment, and the blocks or flow diagrams in the figures are not necessarily required to practice the present invention.
The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for apparatus or system embodiments, since they are substantially similar to method embodiments, they are described in relative terms, as long as they are described in partial descriptions of method embodiments. The above-described embodiments of the apparatus and system are merely illustrative, and the units described as separate parts may or may not be physically separate, and the parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the embodiment of the present invention. One of ordinary skill in the art can understand and implement it without inventive effort.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (7)
1. A method of selecting a perovskite battery modification material, comprising:
acquiring modification material information, perovskite precursor characteristics and device performance parameters before and after optimization of the modification material adopted by the existing perovskite battery, taking the modification material characteristics, the perovskite precursor characteristics and the device performance parameters before optimization of the modification material as input independent variables, and taking the device photovoltaic characteristic parameters after optimization of the modification material as output dependent variables, and establishing an initial data characteristic set;
performing dimensionality reduction on the initial data feature set to obtain an important feature data set;
establishing different machine learning algorithm models, training and testing the different machine learning algorithm models by using the important characteristic data set, evaluating the performance of each machine learning algorithm model, and selecting an optimized machine learning algorithm model;
analyzing and screening key features of the modified materials in the important feature data set by using an optimized machine learning algorithm model, combining the key features to construct the modified materials, predicting performance parameters of the constructed modified materials by using the optimized machine learning algorithm model, and establishing a mapping model between the key features and the final device performance of the perovskite battery by using a machine learning method;
and selecting a modification material according to the perovskite battery device performance required to be realized by utilizing the mapping model.
2. The method of claim 1, wherein the perovskite cell comprises a perovskite solar cell.
3. The method according to claim 1 or 2, wherein the modified material information, perovskite precursor characteristics and device performance parameters before and after optimization of the modified material adopted by the existing perovskite battery are obtained, the modified material characteristics, the perovskite precursor characteristics and the device performance parameters before optimization of the modified material are used as input independent variables, the device photovoltaic characteristic parameters after optimization of the modified material are used as output dependent variables, and an initial data feature set is established and comprises;
extracting a modification material adopted by the perovskite solar cell in a literature report/experimental report, and obtaining SMILES codes of the modification material by adopting molecular editing software or a webpage;
based on SMILES coding, extracting the characteristics of the modified material, then extracting perovskite precursor characteristics and device performance parameters before and after optimization of the modified material, taking the modified material characteristics, the perovskite precursor characteristics and the device performance parameters before optimization of the modified material as input independent variables, taking the device photovoltaic characteristic parameters after optimization of the modified material as output dependent variables, and establishing an initial data characteristic set.
4. The method according to claim 3, wherein the dimension reduction processing is performed on the initial data feature set to obtain an important feature data set, including;
preprocessing the initial set of data features, the preprocessing comprising: cleaning error data and abnormal value data in the initial data feature set, and supplementing some feature missing values in the initial data feature set to obtain a cleaned data set;
and (3) reducing the dimensions of the cleaned data set by adopting correlation analysis, a principal component analysis method, a self-coding grid method or discrete point analysis to obtain an important characteristic data set.
5. The method of claim 4, wherein the establishing different machine learning algorithm models, training and testing the different machine learning algorithm models using the significant feature data set, evaluating performance of each machine learning algorithm model, selecting an optimized machine learning algorithm model, comprises;
establishing different machine learning algorithm models by adopting linear regression, random forests, extreme gradient promotion and neural network algorithms;
dividing the important feature data set into a training set and a testing set, training and testing different machine learning algorithm models by using the training set and the testing set, evaluating the performance of each machine learning algorithm model by using a Root Mean Square Error (RMSE) and a Pearson correlation coefficient r, screening the machine learning algorithm model which executes the lowest RMSE and the highest r value on the testing set, and performing subsequent prediction by using the machine learning algorithm model as an optimized machine learning algorithm model;
the index formula for calculating RMSE is as follows:
the formula for calculating the Pearson correlation coefficient r is as follows:
wherein X i ,Y i ,
And n represents the ith value of the experimental data set, the ith value of the predicted data set, the average value of the experimental data set, and the average value of the predicted data set, respectivelyValues and number of experimental data set points.
6. The method according to claim 5, wherein the optimized machine learning algorithm model is used for analyzing and screening key features of the modified materials in the important feature data set, the key features are combined to construct the modified materials, and the optimized machine learning algorithm model is used for predicting performance parameters of the constructed modified materials, including;
analyzing the importance of the screened optimized machine learning algorithm model, analyzing and quantifying the influence of each characteristic in the important characteristic data set on the output by weight, and screening the most important key characteristic of the modified material;
combining the screened key characteristics to construct a modified material of the perovskite battery, obtaining a SMILES code of the constructed modified material by using molecular editing software or a webpage, extracting the characteristics of the constructed modified material based on the SMILES code, and establishing a prediction data characteristic set based on different perovskite precursor characteristics and device performance parameters before optimization of different modified materials;
and predicting the prediction data characteristic set by adopting an optimized machine learning algorithm to obtain corresponding output values of the modified material of the perovskite battery under different perovskite precursor characteristics and device performance parameters before optimization of different modified materials, wherein the output values comprise device performance parameter values or perovskite characteristic parameter values, and establishing a mapping model between the key characteristics and the final device performance of the perovskite battery by utilizing a machine learning method.
7. The method of claim 1, further comprising;
and carrying out corresponding experimental verification on the perovskite battery modified material selected according to the prediction result of the optimized machine learning algorithm model, carrying out corresponding theoretical verification by adopting device simulation, first principle calculation and related theories, and evaluating the accuracy of the optimized machine learning algorithm model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210569200.3A CN114974476A (en) | 2022-05-24 | 2022-05-24 | Method for selecting perovskite battery modification material |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210569200.3A CN114974476A (en) | 2022-05-24 | 2022-05-24 | Method for selecting perovskite battery modification material |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114974476A true CN114974476A (en) | 2022-08-30 |
Family
ID=82956538
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210569200.3A Pending CN114974476A (en) | 2022-05-24 | 2022-05-24 | Method for selecting perovskite battery modification material |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114974476A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117114143A (en) * | 2023-10-23 | 2023-11-24 | 桑若(厦门)光伏产业有限公司 | Perovskite passivating agent passivation strategy evaluation method, device, equipment and medium |
-
2022
- 2022-05-24 CN CN202210569200.3A patent/CN114974476A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117114143A (en) * | 2023-10-23 | 2023-11-24 | 桑若(厦门)光伏产业有限公司 | Perovskite passivating agent passivation strategy evaluation method, device, equipment and medium |
| CN117114143B (en) * | 2023-10-23 | 2024-02-02 | 桑若(厦门)光伏产业有限公司 | Perovskite passivating agent passivation strategy evaluation method, device, equipment and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109587713B (en) | Network index prediction method and device based on ARIMA model and storage medium | |
| CN109873610B (en) | Photovoltaic array fault diagnosis method based on IV characteristic and depth residual error network | |
| CN110930198A (en) | Electric energy substitution potential prediction method and system based on random forest, storage medium and computer equipment | |
| CN109492748B (en) | Method for establishing medium-and-long-term load prediction model of power system based on convolutional neural network | |
| CN110110915A (en) | A kind of integrated prediction technique of the load based on CNN-SVR model | |
| CN114974476A (en) | Method for selecting perovskite battery modification material | |
| CN110880044B (en) | Markov chain-based load prediction method | |
| CN116842337A (en) | Transformer fault diagnosis method based on LightGBM (gallium nitride based) optimal characteristics and COA-CNN (chip on board) model | |
| CN109921462B (en) | New energy consumption capability assessment method and system based on LSTM | |
| CN111027841A (en) | Low-voltage transformer area line loss calculation method based on gradient lifting decision tree | |
| CN114091776A (en) | K-means-based multi-branch AGCNN short-term power load prediction method | |
| CN114021425A (en) | Power system operation data modeling and feature selection method and device, electronic equipment and storage medium | |
| CN116521495A (en) | System performance bottleneck detection method based on reinforcement learning | |
| CN116191401A (en) | Multi-element load prediction method and system for short-term comprehensive energy system | |
| CN116470491A (en) | Photovoltaic power probability prediction method and system based on copula function | |
| CN115238854A (en) | Short-term load prediction method based on TCN-LSTM-AM | |
| CN113052060B (en) | Bearing residual life prediction method and device based on data enhancement and electronic equipment | |
| CN115392560A (en) | Converter fault prediction method and system based on decision tree | |
| CN114372640A (en) | Wind power prediction method based on fluctuation sequence classification correction | |
| CN114281846A (en) | New energy power generation prediction method based on machine learning | |
| CN113723497A (en) | Abnormal electricity utilization detection method, device, equipment and storage medium based on mixed feature extraction and Stacking model | |
| CN113837486B (en) | RNN-RBM-based distribution network feeder long-term load prediction method | |
| CN112230083B (en) | Method and system for identifying abnormal events of gateway metering device | |
| CN117407681B (en) | Time sequence data prediction model establishment method based on vector clustering | |
| CN116523540A (en) | Ultra-short-term electricity price prediction method based on self-adaptive LGBM |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |