Lithium ion battery multi-metal oxide positive electrode material screening method and system

Abstract

According to the method and the system for screening the multi-metal oxide positive electrode material of the lithium ion battery based on the neural network provided by the invention, the database of the multi-metal oxide positive electrode material of the lithium ion battery is constructed according to the multi-element feature descriptor of the multi-metal oxide positive electrode material of the lithium ion battery; performing deep neural network prediction model training according to the preprocessed database of the multi-metal oxide positive electrode material of the lithium ion battery, and evaluating the trained deep neural network prediction model of the multi-metal oxide positive electrode material of the lithium ion battery. And performing performance prediction and screening on the lithium ion battery multi-metal oxide positive electrode material by using the evaluated deep neural network model of the lithium ion battery multi-metal oxide positive electrode material, the deep neural network having a nonlinear modeling capability and a capability of automatically learning distinctive features. And the complex nonlinear relationship between the characteristics and the performance of the metal positive electrode material can be captured.

Description

Technical Field

[0001] The present application relates to the technical field of material screening, and in particular to a method and system for screening multi-metal oxide positive electrode materials for lithium-ion batteries. Background Art

[0002] With the popularity of electronic devices such as electric vehicles and smart phones, lithium-ion batteries, as rechargeable batteries with high energy density, long cycle life and low self-discharge rate, have become the mainstream energy supply method. Among them, metal oxide positive electrode materials are one of the most important components of lithium-ion batteries, and their performance directly affects the energy density, cycle life and safety of the battery. At present, there have been many studies on metal oxide positive electrode materials, including traditional trial and error methods and first principle calculation methods, as well as machine learning methods that have gradually emerged in recent years.

[0003] The traditional trial and error method usually uses experimental methods to continuously adjust the structure and composition of the material to find the best positive electrode material. The disadvantage of this method is that it is time-consuming, requires a lot of trial and error experiments, and cannot conduct in-depth research on the microstructure of the material. The first principle calculation simulates the structure and properties of the material through a calculation method based on quantum mechanics to predict the performance of the material. The advantage of this method is that it can conduct in-depth research on the microstructure and embedding mechanism of the material, but due to the limitations of calculation accuracy and speed, its application scope has certain limitations. The machine learning method is to establish a prediction model by learning a large amount of sample data to predict the performance of the material. The advantage of this method is that it can process a large amount of data, can simultaneously predict multiple properties of the material, and has a high prediction accuracy. Therefore, the application of machine learning methods in the field of materials science is becoming more and more extensive.

[0004] At present, a prediction model is established through machine learning algorithms to predict the performance of metal oxide cathode materials. The specific method is to first obtain a large number of material descriptors through first-principles calculations, and then use these descriptors as input and the material performance as output to establish a prediction model. Finally, the prediction performance of the model is evaluated through verification of the validation set and the test set. The advantage of this method is that it can process a large amount of data, can predict multiple properties of the material simultaneously, and has a high prediction accuracy. However, since the model needs to rely on the descriptors obtained by first-principles calculations, its scope of application is still limited to a certain extent. Summary of the invention

[0005] In view of this, it is necessary to provide a screening method, screening system, electronic device and computer-readable storage medium for lithium-ion battery multi-metal oxide positive electrode materials that can realize rapid material screening and prediction in response to existing technical defects. As a nonlinear model, deep neural network has great advantages in processing nonlinear relationships of complex data. The prediction ability of complex models can be improved by establishing multiple hidden layers and nonlinear activation functions. At the same time, when faced with multi-scale descriptors such as electronics and structures of metal oxide positive electrode materials, there is no need to manually extract features. The most discriminative features can be automatically learned through multi-layer neural networks, thereby improving the accuracy and generalization of the prediction model of lithium-ion battery metal oxide positive electrode materials.

[0006] To solve the above problems, this application adopts the following technical solutions:

[0007] One of the purposes of the present application is to provide a method for screening multi-metal oxide positive electrode materials for lithium-ion batteries, comprising the following steps:

[0008] Constructing a lithium-ion battery multi-metal oxide positive electrode material database according to the lithium-ion battery multi-metal oxide positive electrode material, wherein the lithium-ion battery multi-metal oxide positive electrode material database includes structural feature data and target data, wherein the feature data includes but is not limited to lattice parameters, bonding information, element composition, unpaired electrons, d-band center, and net charge number, and the target data includes metal vacancy formation energy and oxygen vacancy formation energy;

[0009] Preprocessing the missing value, error value and principal component analysis data of the lithium-ion battery multi-metal oxide positive electrode material database;

[0010] According to the preprocessed lithium-ion battery multi-metal oxide positive electrode material database, a deep neural network prediction model is used for training;

[0011] Evaluate the accuracy and generalization of trained neural network models for multi-metal oxide cathode materials for lithium-ion batteries;

[0012] The evaluated multi-metal oxide positive electrode material model for lithium-ion batteries is used to predict and screen multi-metal oxide positive electrode materials for lithium-ion batteries.

[0013] In some of the embodiments, the step of constructing a lithium-ion battery multi-metal oxide positive electrode material database according to the lithium-ion battery multi-metal oxide positive electrode material specifically includes the following steps:

[0014] Obtaining existing lithium-ion battery multi-metal oxide positive electrode materials and extracting their structural feature data and using principal component analysis combined with a neural network algorithm to identify highly distinguishing features, wherein the lithium-ion battery multi-metal oxide positive electrode materials include but are not limited to transition metals, wherein the transition metals include but are not limited to manganese-based, cobalt-based and iron-based, and the feature data include but are not limited to lattice parameters, bonding information, elemental composition, unpaired electrons, d-band centers, and net charge numbers;

[0015] Calculating the metal vacancy formation energy and oxygen vacancy formation energy of some structures in the multi-metal oxide positive electrode material of the lithium-ion battery and the metal vacancy formation energy and oxygen vacancy formation energy of different lithium insertion concentrations, and taking the metal vacancy formation energy and oxygen vacancy formation energy as target data;

[0016] The characteristic data and the target data are incorporated into the multi-metal oxide positive electrode database.

[0017] In some of the embodiments, the steps of obtaining the existing lithium-ion battery multi-metal oxide positive electrode material and extracting its structural characteristics specifically include the following steps:

[0018] Based on crawler technology and structural database interface, the existing multi-metal oxide positive electrode material structure of lithium-ion batteries is searched, the first principle calculation method is used to optimize part of the structure of the multi-metal oxide positive electrode material of lithium-ion batteries, and the optimized structure is statically calculated to collect structural feature data.

[0019] In some of the embodiments, the step of calculating the metal vacancy formation energy and oxygen vacancy formation energy of some structures in the multi-metal oxide positive electrode material of the lithium ion battery and the metal vacancy formation energy and oxygen vacancy formation energy of different lithium insertion concentrations specifically includes the following steps:

[0020] Based on the first principle calculation, the metal vacancy formation energy and oxygen vacancy formation energy of some structures in the lithium-ion battery multi-metal oxide positive electrode material as well as the metal vacancy formation energy and oxygen vacancy formation energy of different lithium insertion concentrations are calculated.

[0021] In some embodiments, the step of preprocessing the data of the lithium-ion battery multi-metal oxide positive electrode material database specifically includes the following steps:

[0022] The data in the lithium-ion battery multi-metal oxide positive electrode material database are preprocessed, and the intrinsic characteristic parameters of the lithium-ion battery multi-metal oxide positive electrode material are added to the lithium-ion battery multi-metal oxide positive electrode material database and principal component analysis is performed, and the intrinsic characteristic parameters with low correlation are selected as input data, the preprocessing includes data cleaning, removal of outliers, data normalization and feature selection, and the intrinsic characteristic parameters include but are not limited to the electronegativity, ionization energy, magnetic moment, number of valence electrons and unpaired electrons of the metal components in the lithium-ion battery multi-metal oxide positive electrode material.

[0023] In some embodiments, the step of training a prediction model based on the preprocessed lithium-ion battery multi-metal oxide positive electrode material database specifically includes the following steps:

[0024] The preprocessed lithium-ion battery multi-metal oxide positive electrode material database is divided into a training set and a test set, and a deep neural network algorithm is used to train the model. The accuracy and generalization of the model are improved by adjusting the hidden layer and the nonlinear activation function.

[0025] In some embodiments, the step of evaluating the trained lithium-ion battery multi-metal oxide positive electrode material model specifically includes the following steps:

[0026] Cross-validation is used to evaluate the trained lithium-ion battery multi-metal oxide positive electrode material model, and the evaluation indicators include mean square error, determination coefficient and mean absolute error.

[0027] In some of the embodiments, the step of screening the multi-metal oxide positive electrode material for lithium-ion batteries using the evaluated multi-metal oxide positive electrode material model for lithium-ion batteries specifically includes the following steps:

[0028] By inputting the characteristic data of the new material into the lithium-ion battery multi-metal oxide positive electrode material model, the performance prediction result of the new material can be obtained.

[0029] The second object of the present application is to provide a screening system for the screening method of the lithium-ion battery multi-metal oxide positive electrode material, comprising:

[0030] A database construction unit, for constructing a lithium-ion battery multi-metal oxide positive electrode material database according to the lithium-ion battery multi-metal oxide positive electrode material, wherein the lithium-ion battery multi-metal oxide positive electrode material database includes structural feature data and target data, wherein the feature data includes but is not limited to lattice parameters, bonding information, element composition, unpaired electrons, d-band center, net charge number, and the target data includes metal vacancy formation energy and oxygen vacancy formation energy;

[0031] A processing unit, used for performing data preprocessing on the lithium-ion battery multi-metal oxide positive electrode material database;

[0032] A training unit, used for performing prediction model training according to the preprocessed lithium-ion battery multi-metal oxide positive electrode material database;

[0033] An evaluation unit, used to evaluate the trained multi-metal oxide cathode material model for lithium-ion batteries;

[0034] A screening unit is used to screen the multi-metal oxide positive electrode material for lithium-ion batteries using the evaluated multi-metal oxide positive electrode material model for lithium-ion batteries.

[0035] This application adopts the above technical solution, and its beneficial effects are as follows:

[0036] The present application provides a screening method and screening system for lithium-ion battery multi-metal oxide positive electrode materials, a lithium-ion battery multi-metal oxide positive electrode material database is constructed based on the lithium-ion battery multi-metal oxide positive electrode material, data preprocessing is performed on the lithium-ion battery multi-metal oxide positive electrode material database, prediction model training is performed based on the preprocessed lithium-ion battery multi-metal oxide positive electrode material database, the trained lithium-ion battery multi-metal oxide positive electrode material model is evaluated, and the lithium-ion battery multi-metal oxide positive electrode material model after evaluation is used to screen the lithium-ion battery multi-metal oxide positive electrode material, effectively utilizing the high-throughput calculation of the computer and the advanced machine learning algorithm, obtaining part of the data set through high-throughput calculation at low cost, and directly predicting the performance of hundreds of lithium-ion battery multi-metal oxide positive electrode materials through the data-oriented mathematical fitting method of the machine learning algorithm. Compared with the current trial and error method at the experimental level, it is more efficient and less costly; compared with the high-throughput calculation at the purely theoretical level, a large amount of machine time and resource costs are reduced, and only part of the computing resources need to be used to build the initial database. The method has a clear purpose, high efficiency, and saves experimental consumables and computing resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments of the present application or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0038] Figure 1 This is a flow chart of the steps of the method for screening multi-metal oxide positive electrode materials for lithium-ion batteries provided in Example 1 of the present invention.

[0039] Figure 2Schematic diagram of the principle of the method for screening multi-metal oxide positive electrode materials for lithium-ion batteries provided in Example 1 of the present invention.

[0040] Figure 3 A schematic structural diagram of the screening of multi-metal oxide positive electrode materials for lithium-ion batteries provided in Example 2 of the present invention. DETAILED DESCRIPTION

[0041] Embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0042] In the description of the present application, it should be understood that the terms "upper", "lower", "horizontal", "inside", "outside", etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present application.

[0043] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of this application, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0044] In order to make the objectives, technical solutions and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments.

[0045] Example 1

[0046] See also Figure 1 and Figure 2 , respectively, are a step flow chart and a principle schematic diagram of the screening method for multi-metal oxide positive electrode materials for lithium-ion batteries provided in this embodiment, including the following steps S110 to S150, and the specific implementation method of each step is described in detail below.

[0047] Step S110: constructing a lithium-ion battery multi-metal oxide positive electrode material database based on the lithium-ion battery multi-metal oxide positive electrode material, wherein the lithium-ion battery multi-metal oxide positive electrode material database includes structural feature data and target data, wherein the feature data includes but is not limited to lattice parameters, bonding information, element composition, unpaired electrons, d-band center, and net charge number, and the target data includes metal vacancy formation energy and oxygen vacancy formation energy.

[0048] In this embodiment, the step of constructing a lithium-ion battery multi-metal oxide positive electrode material database according to lithium-ion battery multi-metal oxide positive electrode materials specifically includes the following steps S111 to S113, and the specific implementation method of each step is described in detail below.

[0049] Step S111: obtaining an existing lithium-ion battery multi-metal oxide positive electrode material and extracting its structural feature data.

[0050] In this embodiment, the lithium-ion battery multi-metal oxide positive electrode material includes but is not limited to transition metals, and the transition metals include manganese-based, cobalt-based and iron-based. The characteristic data include but are not limited to lattice parameters, bonding information, element composition, unpaired electrons, d-band center, and net charge number.

[0051] It can be understood that the present embodiment proposes more accurate descriptors including the number of unpaired electrons, net charge and d-band center, which have been proven to be accurate structural stability descriptors, effectively screening the framework materials and active components of metal oxides, and achieving efficient screening and prediction of high-performance lithium-ion battery multi-metal oxide positive electrode materials.

[0052] Specifically, based on crawler technology and structural database interface, the existing multi-metal oxide positive electrode material structure of lithium-ion batteries is searched, the first-principles calculation method is used to optimize part of the structure of the multi-metal oxide positive electrode material of lithium-ion batteries, and the optimized structure is statically calculated to collect structural feature data.

[0053] Step S112: calculating the metal vacancy formation energy and oxygen vacancy formation energy of some structures in the lithium-ion battery multi-metal oxide positive electrode material and the metal vacancy formation energy and oxygen vacancy formation energy of different lithium insertion concentrations, and taking the metal vacancy formation energy and oxygen vacancy formation energy as target data.

[0054] Specifically, the metal vacancy formation energy and oxygen vacancy formation energy of some structures in the lithium-ion battery multi-metal oxide positive electrode material as well as the metal vacancy formation energy and oxygen vacancy formation energy of different lithium insertion concentrations are calculated based on the first principle calculation.

[0055] Step S113: Incorporating the characteristic data and the target data into the multi-metal oxide positive electrode database.

[0056] It can be understood that this embodiment calculates the metal vacancy formation energy and oxygen vacancy formation energy at a high throughput based on the multi-metal oxide bonding level and for multi-metal oxide structures with different lithium ion embedding concentrations, thereby screening out stable structures to resist the dissolution of metals and oxygen during the charge and discharge process of lithium-ion batteries, thereby improving the accuracy of screening.

[0057] Step S120: performing data preprocessing on the lithium-ion battery multi-metal oxide positive electrode material database.

[0058] In this embodiment, the step of preprocessing the missing value, error value and principal component analysis data of the lithium-ion battery multi-metal oxide positive electrode material database specifically includes the following steps:

[0059] The data in the lithium-ion battery multi-metal oxide positive electrode material database are preprocessed, and the intrinsic characteristic parameters of the lithium-ion battery multi-metal oxide positive electrode material are added to the lithium-ion battery multi-metal oxide positive electrode material database and a correlation analysis is performed, and the intrinsic characteristic parameters with low correlation are selected as input data, the preprocessing includes data cleaning, removal of outliers, data normalization and feature selection, and the intrinsic characteristic parameters include the electronegativity, ionization energy, magnetic moment, number of valence electrons and unpaired electrons of the metal components in the lithium-ion battery multi-metal oxide positive electrode material.

[0060] It can be understood that this embodiment includes the three key characteristic values ​​of unpaired electrons, net charge number and magnetic moment into the database category. Compared with basic electronic state information such as electronegativity and valence electron number, the key characteristic values ​​adopted in this scheme are more accurate.

[0061] Step S130: Based on the preprocessed multi-metal oxide positive electrode material database for lithium-ion batteries, a deep neural network is used to train a prediction model. In this embodiment, the step of training a prediction model based on the preprocessed multi-metal oxide positive electrode material database for lithium-ion batteries specifically includes the following steps: dividing the preprocessed multi-metal oxide positive electrode material database for lithium-ion batteries into a training set and a validation set, and using a machine learning algorithm to train the model.

[0062] It can be understood that this embodiment divides the preprocessed data set into a training set and a validation set, uses a machine learning algorithm to train the model, and selects a regression algorithm suitable for processing multi-feature data, such as a support vector machine, a neural network, etc. During the model training process, the model hyperparameters can be adjusted, appropriate features with normal data distribution can be selected, and ensemble learning and other methods can be used to improve the prediction performance of the model.

[0063] Step S140: Evaluate the accuracy and generalization of the trained neural network model of lithium-ion battery multi-metal oxide positive electrode materials.

[0064] In this embodiment, the step of evaluating the trained lithium-ion battery multi-metal oxide positive electrode material model specifically includes the following steps: cross-validation is used to evaluate the trained lithium-ion battery multi-metal oxide positive electrode material model, and the evaluation indicators include mean square error, determination coefficient and mean absolute error.

[0065] It can be understood that this embodiment uses cross-validation and other methods to evaluate the trained model to measure the prediction accuracy and generalization ability of the model. If the evaluation effect is not ideal, the model hyperparameters are adjusted and the above step S110 is repeated to include more metal oxide types and information to improve the accuracy and generalization ability of the model until the model is accurate and usable.

[0066] Step S150: using the evaluated lithium-ion battery multi-metal oxide positive electrode material model to predict and screen lithium-ion battery multi-metal oxide positive electrode materials.

[0067] In this embodiment, the step of screening the multi-metal oxide positive electrode materials for lithium-ion batteries using the evaluated multi-metal oxide positive electrode material model for lithium-ion batteries specifically includes the following steps: inputting the characteristic data of the new material into the multi-metal oxide positive electrode material model for lithium-ion batteries to obtain the performance prediction result of the new material.

[0068] It can be understood that the optimized prediction model is applied to the screening of multi-metal oxide positive electrode materials for lithium-ion batteries, and the characteristic data of new materials are input into the model to obtain the performance prediction results of new materials. Through the interpretation of the model, the structure and properties of new materials can be understood, providing guidance for subsequent experimental design and optimization. At the same time, the metal vacancy formation energy and oxygen vacancy formation energy of the predicted material can be calculated and verified by the first principle calculation. If they are basically consistent, it means that the model is accurate. If not, repeat the above steps to incorporate the properties of the new element into the database.

[0069] It can be understood that the above embodiments of the present application construct a data set of characteristic parameters of metal oxygen bonds in multi-metal oxides through first-principle calculations, and use electronic structure and geometric structure as characteristic parameters, metal vacancy formation energy and oxygen vacancy formation energy as structural stability evaluation indicators, and combine multiple machine learning algorithms to screen and predict stable structures, thereby realizing rapid screening and prediction of stable structures of multi-metal oxide positive electrode materials for lithium-ion batteries.

[0070] The screening method for lithium-ion battery multi-metal oxide positive electrode materials provided in the above embodiments of the present application effectively utilizes the high-throughput computing and advanced machine learning algorithms of computers, obtains partial data sets through high-throughput computing at low cost, and directly predicts the performance of hundreds of lithium-ion battery multi-metal oxide positive electrode materials through machine learning algorithms, which are data-oriented mathematical fitting methods. Compared with the current trial-and-error method at the experimental level, it is more efficient and less costly; compared with high-throughput computing at the purely theoretical level, it reduces a lot of machine time and resource costs, and only needs to use part of the computing resources to build the initial database. This method is purposeful, efficient, and saves experimental consumables and computing resources.

[0071] Example 2

[0072] See also Figure 3 , which is a structural schematic diagram of a screening system for multi-metal oxide positive electrode materials for lithium-ion batteries provided in this embodiment, comprising a database construction unit 110, for constructing a multi-metal oxide positive electrode material database for lithium-ion batteries according to the multi-metal oxide positive electrode materials for lithium-ion batteries; a processing unit 120, for performing data preprocessing on the multi-metal oxide positive electrode material database for lithium-ion batteries; a training unit 130, for performing prediction model training on the multi-metal oxide positive electrode material database for lithium-ion batteries after preprocessing; an evaluation unit 140, for evaluating the trained multi-metal oxide positive electrode material model for lithium-ion batteries; and a screening unit 150, for screening multi-metal oxide positive electrode materials for lithium-ion batteries using the evaluated multi-metal oxide positive electrode material model for lithium-ion batteries.

[0073] The screening system for multi-metal oxide positive electrode materials for lithium-ion batteries provided in this embodiment can be implemented in detail by referring to Embodiment 1, which will not be described in detail here.

[0074] The screening system for lithium-ion battery multi-metal oxide positive electrode materials provided in the above-mentioned embodiment 2 of the present application effectively utilizes the high-throughput computing and advanced machine learning algorithms of computers, obtains partial data sets through high-throughput computing at low cost, and directly predicts the performance of hundreds of lithium-ion battery multi-metal oxide positive electrode materials through machine learning algorithms, a data-oriented mathematical fitting method. Compared with the current trial-and-error method at the experimental level, it is more efficient and less costly; compared with the high-throughput computing at the purely theoretical level, it reduces a lot of machine time and resource costs, and only needs to use part of the computing resources to build the initial database. This method is purposeful, efficient, and saves experimental consumables and computing resources.

[0075] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present application.

[0076] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0077] It can be understood that the technical features of the above-described embodiments can be arbitrarily combined. In order to make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0078] The above are only preferred embodiments of the present application, and only specifically describe the technical principles of the present application. These descriptions are only for explaining the principles of the present application and cannot be interpreted as limiting the scope of protection of the present application in any way. Based on the explanation here, any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application, and other specific implementation methods of the present application that can be associated with the technicians in this field without creative work, should be included in the scope of protection of the present application.

Claims

1. A method for screening multi-metal oxide positive electrode materials for lithium-ion batteries, characterized in that: The steps include: Constructing a lithium-ion battery multi-metal oxide positive electrode material database according to the lithium-ion battery multi-metal oxide positive electrode material, wherein the lithium-ion battery multi-metal oxide positive electrode material database includes structural feature data and target data, wherein the feature data includes but is not limited to lattice parameters, bonding information, element composition, unpaired electrons, d-band center, and net charge number, and the target data includes metal vacancy formation energy and oxygen vacancy formation energy; Preprocessing the missing value, error value and principal component analysis data of the lithium-ion battery multi-metal oxide positive electrode material database; According to the preprocessed lithium-ion battery multi-metal oxide positive electrode material database, a deep neural network prediction model is used for training; Evaluate the accuracy and generalization of trained neural network models for multi-metal oxide cathode materials for lithium-ion batteries; The evaluated multi-metal oxide positive electrode material model for lithium-ion batteries is used to predict and screen multi-metal oxide positive electrode materials for lithium-ion batteries.

2. The method for screening multi-metal oxide positive electrode materials for lithium-ion batteries according to claim 1, characterized in that: The step of constructing a lithium-ion battery multi-metal oxide positive electrode material database according to the lithium-ion battery multi-metal oxide positive electrode material specifically includes the following steps: Obtaining existing lithium-ion battery multi-metal oxide positive electrode materials and extracting their structural feature data and using principal component analysis combined with a neural network algorithm to identify highly distinguishing features, wherein the lithium-ion battery multi-metal oxide positive electrode materials include but are not limited to transition metals, and the transition metals include manganese-based, cobalt-based and iron-based, and the feature data include but are not limited to lattice parameters, bonding information, elemental composition, unpaired electrons, d-band centers, and net charge numbers; Calculating the metal vacancy formation energy and oxygen vacancy formation energy of some structures in the multi-metal oxide positive electrode material of the lithium-ion battery and the metal vacancy formation energy and oxygen vacancy formation energy of different lithium insertion concentrations, and taking the metal vacancy formation energy and oxygen vacancy formation energy as target data; The characteristic data and the target data are incorporated into the multi-metal oxide positive electrode database.

3. The method for screening multi-metal oxide positive electrode materials for lithium-ion batteries according to claim 2, characterized in that: The steps of obtaining the existing lithium-ion battery multi-metal oxide positive electrode material and extracting its structural characteristics specifically include the following steps: Based on crawler technology and structural database interface, the existing multi-metal oxide positive electrode material structure of lithium-ion batteries is searched, the first-principles calculation method is used to optimize part of the structure of the multi-metal oxide positive electrode material of lithium-ion batteries, and the optimized structure is statically calculated to collect structural feature data.

4. The method for screening multi-metal oxide positive electrode materials for lithium-ion batteries according to claim 2, characterized in that: The step of calculating the metal vacancy formation energy and oxygen vacancy formation energy of some structures in the multi-metal oxide positive electrode material for lithium-ion batteries and the metal vacancy formation energy and oxygen vacancy formation energy of different lithium insertion concentrations specifically includes the following steps: Based on the first principle calculation, the metal vacancy formation energy and oxygen vacancy formation energy of some structures in the lithium-ion battery multi-metal oxide positive electrode material as well as the metal vacancy formation energy and oxygen vacancy formation energy of different lithium insertion concentrations are calculated.

5. The method for screening multi-metal oxide positive electrode materials for lithium-ion batteries according to claim 2, characterized in that: The step of preprocessing the data of the lithium-ion battery multi-metal oxide positive electrode material database specifically includes the following steps: The data in the lithium-ion battery multi-metal oxide positive electrode material database are preprocessed, and the intrinsic characteristic parameters of the lithium-ion battery multi-metal oxide positive electrode material are added to the lithium-ion battery multi-metal oxide positive electrode material database, and principal component analysis is performed, and the intrinsic characteristic parameters with low correlation are selected as input data, the preprocessing includes data cleaning, removal of outliers, data normalization and feature selection, and the intrinsic characteristic parameters include the electronegativity, ionization energy, magnetic moment, number of valence electrons and unpaired electrons of the metal components in the lithium-ion battery multi-metal oxide positive electrode material.

6. The method for screening multi-metal oxide positive electrode materials for lithium-ion batteries according to claim 5, characterized in that: The step of training a prediction model according to the preprocessed lithium-ion battery multi-metal oxide positive electrode material database specifically includes the following steps: The preprocessed lithium-ion battery multi-metal oxide positive electrode material database is divided into a training set and a test set, and the model is trained using a deep neural network machine learning algorithm. The hidden layer and nonlinear activation function are adjusted to improve the model accuracy and generalization.

7. The method for screening multi-metal oxide positive electrode materials for lithium-ion batteries according to claim 6, characterized in that: The steps of evaluating the trained lithium-ion battery multi-metal oxide positive electrode material model specifically include the following steps: Cross-validation is used to evaluate the trained lithium-ion battery multi-metal oxide positive electrode material model, and the evaluation indicators include mean square error, determination coefficient and mean absolute error.

8. The method for screening multi-metal oxide positive electrode materials for lithium-ion batteries according to claim 7, characterized in that: The step of screening the multi-metal oxide positive electrode material for lithium-ion batteries using the evaluated multi-metal oxide positive electrode material model for lithium-ion batteries specifically includes the following steps: By inputting the characteristic data of the new material into the neural network model of the lithium-ion battery multi-metal oxide positive electrode material, the performance prediction result of the new material can be obtained.

9. A screening system for the screening method of lithium-ion battery multi-metal oxide positive electrode materials as claimed in claim 1, characterized in that: include: A database construction unit, for constructing a lithium-ion battery multi-metal oxide positive electrode material database according to the lithium-ion battery multi-metal oxide positive electrode material, wherein the lithium-ion battery multi-metal oxide positive electrode material database includes structural feature data and target data, wherein the feature data includes but is not limited to lattice parameters, bonding information, element composition, unpaired electrons, d-band center, net charge number, and the target data includes metal vacancy formation energy and oxygen vacancy formation energy; A processing unit, used for performing data preprocessing on the lithium-ion battery multi-metal oxide positive electrode material database; A training unit, used for training a neural network prediction model according to the preprocessed lithium-ion battery multi-metal oxide positive electrode material database; An evaluation unit, used to evaluate the trained neural network model of multi-metal oxide cathode materials for lithium-ion batteries; A screening unit is used to screen the multi-metal oxide positive electrode material for lithium-ion batteries using the evaluated neural network model of the multi-metal oxide positive electrode material for lithium-ion batteries.

Images