Component optimization method and system for ultra-high performance lithium battery
By constructing a coding space and morphology prediction model, the component configuration of ultra-high-performance lithium batteries is optimized, the problem of low efficiency of component configuration is solved, and efficient adaptation of batteries and scenarios is achieved.
Patent Information
- Application Number
- CN202510731210.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-03
- Publication Date
- 2025-09-12
AI Technical Summary
The existing technology has low efficiency in configuring the components of ultra-high-performance lithium batteries, making it difficult to quickly and accurately meet the performance requirements of different application scenarios, affecting the compatibility of the battery with the scenario.
By constructing the nickel content coding space and the doping element coding space, combined with the nickel-rich layered transition metal oxide morphology prediction model, random coding and morphology prediction are performed to screen out component combinations with high similarity to the standard morphology and return them to the client for experimental verification.
It significantly improves the efficiency of component configuration, enhances the adaptability of batteries to application scenarios, and reduces experimental blindness and resource consumption.
Smart Images
Figure CN120636593A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of intelligent manufacturing technology, and in particular to a composition optimization method and system for an ultra-high performance lithium battery. Background Art
[0002] New energy vehicles, portable electronic devices, and energy storage systems are all important application scenarios for ultra-high-performance lithium batteries. Depending on different needs, the performance requirements of ultra-high-performance lithium batteries are also different. The battery composition scheme needs to be configured in the early stage of battery preparation. At present, in the process of preparing ultra-high-performance lithium batteries, the composition configuration of nickel-rich layered transition metal oxides is usually based on existing experience and is explored and determined through a large number of experiments. This method has certain limitations in practical applications.
[0003] Traditional methods for determining the composition of ultra-high-performance lithium-ion batteries require extensive experimentation, which is inefficient. Furthermore, due to the complexity and uncertainty of these experiments, it is difficult to quickly and accurately identify a compositional configuration that meets the performance requirements for different application scenarios, making it difficult to adapt to the growing demand for ultra-high-performance lithium-ion batteries. This leads to technical issues such as low compositional efficiency and poor compatibility between batteries and specific scenarios. Summary of the Invention
[0004] The present invention provides a composition optimization method and system for ultra-high-performance lithium batteries to solve the technical problems in the prior art of low composition configuration efficiency and impact on battery and scene adaptability, thereby achieving the technical effects of improved composition configuration efficiency and optimized battery and scene adaptability.
[0005] In a first aspect, the present invention provides a method for optimizing the composition of an ultra-high performance lithium battery, wherein the method comprises:
[0006] Receive from the client the nickel content constraint range, doping element constraint range and standard morphology of nickel-rich layered transition metal oxides for ultra-high performance lithium batteries.
[0007] According to the nickel content constraint interval and the doping element constraint interval, a nickel content coding space and a doping element coding space are constructed.
[0008] A nickel-rich layered transition metal oxide morphology prediction model is downloaded based on the doping element properties, wherein the nickel-rich layered transition metal oxide morphology prediction model is trained by multiple sets of data, and any set of the multiple sets of data includes: nickel content coded record data, doping element content coded record data and a label identifying the oxide morphology.
[0009] Random encoding is performed according to the nickel content encoding space and the doping element encoding space to obtain component content encoding data, and the component content encoding data is processed by the nickel-rich layered transition metal oxide morphology prediction model to obtain the predicted morphology of the nickel-rich layered transition metal oxide.
[0010] When the morphological similarity between the standard morphology of the nickel-rich layered transition metal oxide and the predicted morphology of the nickel-rich layered transition metal oxide is greater than or equal to a morphological similarity threshold, the component content encoding data is experimentally identified and returned to the client.
[0011] In a second aspect, the present invention further provides a composition optimization system for ultra-high performance lithium batteries, wherein the system comprises:
[0012] The client data receiving component is used to receive the nickel content constraint range, doping element constraint range and standard morphology of the nickel-rich layered transition metal oxide for ultra-high performance lithium batteries from the client.
[0013] The coding space construction component is used to construct the nickel content coding space and the doping element coding space according to the nickel content constraint interval and the doping element constraint interval.
[0014] A morphology prediction model download component is used to download a nickel-rich layered transition metal oxide morphology prediction model based on the properties of doping elements, wherein the nickel-rich layered transition metal oxide morphology prediction model is trained by multiple sets of data, and any set of the multiple sets of data includes: nickel content coded record data, doping element content coded record data and a label identifying the oxide morphology.
[0015] The predicted morphology acquisition component is used to perform random encoding according to the nickel content encoding space and the doping element encoding space to obtain component content encoding data, and process the component content encoding data through the nickel-rich layered transition metal oxide morphology prediction model to obtain the predicted morphology of the nickel-rich layered transition metal oxide.
[0016] The experimental identification returning component is used to return the component content encoding data to the client with experimental identification when the morphological similarity between the standard morphology of the nickel-rich layered transition metal oxide and the predicted morphology of the nickel-rich layered transition metal oxide is greater than or equal to the morphological similarity threshold.
[0017] The present invention discloses a composition optimization method and system for ultra-high-performance lithium batteries, comprising: receiving nickel content constraint intervals, doping element constraint intervals, and standard morphology information of nickel-rich layered transition metal oxides for ultra-high-performance lithium batteries from a client; constructing a nickel content coding space and a doping element coding space based on the nickel content constraint intervals and the doping element constraint intervals; and downloading a nickel-rich layered transition metal oxide morphology prediction model based on the characteristics of the doping element. The model is trained using multiple data sets, each set containing: nickel content coding data, doping element content coding data, and a label identifying the oxide morphology; random encoding is performed within the nickel content coding space and the doping element coding space to obtain component content coding data. These coded data are then processed through a nickel-rich layered transition metal oxide morphology prediction model to obtain a predicted morphology of the nickel-rich layered transition metal oxide; when the similarity between the standard morphology of the nickel-rich layered transition metal oxide and the predicted morphology is greater than or equal to a preset morphology similarity threshold, the component content coded data is identified as experimental data and returned to the client. The present invention discloses a composition optimization method and system for ultra-high performance lithium batteries that solves the technical problems of low component configuration efficiency and impact on battery and scene adaptability, and achieves the technical effects of improved component configuration efficiency and optimized battery and scene adaptability. BRIEF DESCRIPTION OF THE DRAWINGS
[0018] Figure 1 Schematic diagram of a process for optimizing the composition of an ultra-high performance lithium battery according to the present invention;
[0019] Figure 2 This is a structural schematic diagram of a composition optimization system for an ultra-high performance lithium battery according to the present invention.
[0020] Description of the accompanying drawings: client data receiving component 11, encoding space construction component 12, morphology prediction model downloading component 13, predicted morphology acquisition component 14, experiment identifier returning component 15. DETAILED DESCRIPTION
[0021] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts shall fall within the scope of protection of the present invention.
[0022] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature specified as "first" or "second" may explicitly or implicitly include one or more of the specified features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.
[0023] In the description of the present invention, the term "for example" is used to mean "used as an example, illustration or illustration". Any embodiment of the present invention described as "for example" is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any person skilled in the art to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes are not elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed herein.
[0024] Example 1, as Figure 1 , a schematic flow chart of a method for optimizing the composition of an ultra-high performance lithium battery according to the present invention, wherein the method comprises:
[0025] Receive from the client the nickel content constraint range, doping element constraint range and standard morphology of nickel-rich layered transition metal oxides for ultra-high performance lithium batteries.
[0026] Specifically, the client is a terminal device or system operated by a user to interact with the server, and is used to input required parameters and receive results returned by the server.
[0027] Specifically, nickel-rich layered transition metal oxides are key positive electrode materials for ultra-high-performance lithium batteries. Their nickel content and the types and proportions of doping elements have an important impact on battery performance. The nickel content constraint range and the doping element constraint range refer to the ranges of nickel content and doping element content set by users based on battery performance requirements. The standard morphology refers to the microstructural characteristics of nickel-rich layered transition metal oxides under ideal conditions, including particle size, shape, distribution, etc.
[0028] Specifically, the server receives key user input via a communication interface with the client. These parameters include the nickel content range, doping element range, and desired standard morphology of the nickel-rich layered transition metal oxide, as determined by the user's specific battery performance requirements for the application scenario (e.g., new energy vehicles, portable electronic devices, etc.). For example, for new energy vehicles, users may desire higher energy density, thus setting a higher nickel content range; whereas for portable electronic devices, cycle life may be more important, and the choice of doping element may differ.
[0029] The parameters input by the client clarify user needs, provide an optimization basis for subsequent coding space construction and model prediction, and ensure that the optimization process can be customized to specific needs.
[0030] According to the nickel content constraint interval and the doping element constraint interval, a nickel content coding space and a doping element coding space are constructed.
[0031] Specifically, the nickel content coding space and the doping element coding space represent all possible combinations of nickel content and doping element content. Each dimension in the coding space represents the numerical range of a parameter, such as the integer or decimal part of the nickel content or the content of a certain doping element. By constructing the coding space, complex composition parameters can be converted into an operational mathematical model, which facilitates subsequent random coding and optimization processing.
[0032] In some embodiments, constructing a nickel content coding space and a doping element coding space according to the nickel content constraint interval and the doping element constraint interval includes:
[0033] Obtain the maximum number of digits of the integer part and the maximum number of digits of the decimal part of the nickel content constraint interval; construct a multidimensional coding space for the integer part according to the maximum number of digits of the integer part, and construct a multidimensional coding space for the decimal part according to the maximum number of digits of the decimal part, wherein the dimension of the multidimensional coding space for the integer part is equal to the maximum number of digits of the integer part, and the dimension of the multidimensional coding space for the decimal part is equal to the maximum number of digits of the decimal part; extract the constraint interval for each digit of the integer part according to the constraint interval for each digit of the integer part, perform regional restrictions on the multidimensional coding space for the integer part according to the constraint interval for each digit of the integer part, and obtain the coding space for the integer part of the nickel content; extract the constraint interval for each digit of the decimal part according to the constraint interval for each digit of the decimal part, perform regional restrictions on the multidimensional coding space for the decimal part according to the constraint interval for each digit of the decimal part, and obtain the coding space for the decimal part of the nickel content; set the coding space for the integer part of the nickel content and the coding space for the decimal part of the nickel content as the nickel content coding space; wherein the construction process of the doping element coding space is the same as that of the nickel content coding space.
[0034] Specifically, the maximum number of digits in the integer part and the maximum number of digits in the fractional part refer to the maximum number of digits in the integer and fractional parts within the nickel content constraint interval, and are used to determine the dimensionality of the coding space. For example, if the nickel content constraint interval is [0.85, 0.95], the maximum number of digits in the integer part is 0, and the maximum number of digits in the fractional part is 2. A multidimensional coding space is a virtual mathematical space used to represent all possible values of a parameter, where each dimension corresponds to the numerical range of a parameter. By restricting the range of these dimensions, a coding space that meets the constraints can be constructed.
[0035] Specifically, based on the nickel content constraint interval and the doping element constraint interval, a nickel content encoding space and a doping element encoding space are constructed. First, the maximum number of digits in the integer part and the maximum number of digits in the decimal part of the nickel content constraint interval are extracted. For example, if the nickel content constraint interval is [0.85, 0.95], the integer part is 0 and the decimal part is two digits. Then, a multidimensional encoding space for the integer part is constructed based on the maximum number of digits in the integer part, and a multidimensional encoding space for the decimal part is constructed based on the maximum number of digits in the decimal part. For example, for two decimal digits, a two-dimensional space can be constructed, with each dimension representing the first and second digits after the decimal point. Then, based on the constraint interval, the constraint intervals for each digit of the integer part and the decimal part are extracted, and regional restrictions are applied to the multidimensional encoding space, ultimately obtaining the integer part encoding space and the decimal part encoding space of the nickel content. Finally, the integer part encoding space and the decimal part encoding space are merged to form a complete nickel content encoding space.
[0036] The above process transforms complex constraints into an ordered coding space through precise mathematical modeling, providing a clear scope and structure for subsequent random coding and model prediction, ensuring that the component optimization process can be carried out efficiently within the constraints set by the user.
[0037] A nickel-rich layered transition metal oxide morphology prediction model is downloaded based on the doping element properties, wherein the nickel-rich layered transition metal oxide morphology prediction model is trained by multiple sets of data, and any set of the multiple sets of data includes: nickel content coded record data, doping element content coded record data and a label identifying the oxide morphology.
[0038] Specifically, the doping element properties refer to the type of doping elements (such as aluminum, cobalt, manganese, etc.) and their related physical or chemical properties. The doping element properties will significantly affect the micromorphology of nickel-rich layered transition metal oxides; the nickel-rich layered transition metal oxide morphology prediction model is a model constructed based on machine learning or artificial intelligence algorithms, which is used to predict the micromorphology of nickel-rich layered transition metal oxides based on nickel content and doping element content.
[0039] Specifically, the nickel content coded record data and the doping element content coded record data refer to the numerical data of nickel content and doping element content that have been encoded, and these data serve as training inputs for the nickel-rich layered transition metal oxide morphology prediction model; the label identifying the oxide morphology refers to the actual morphological characteristics of the nickel-rich layered transition metal oxide corresponding to the above input data, such as particle size, shape, distribution, etc., which are used as expected outputs during model supervised training.
[0040] In some embodiments, the nickel-rich layered transition metal oxide morphology prediction model training process includes:
[0041] Deploy a trusted third party; after sending a doping element property set and an encryption key of a nickel-rich layered transition metal oxide to several ultra-high performance lithium battery preparation nodes through the trusted third party, receive several feedback information sets encrypted and transmitted back by the several ultra-high performance lithium battery preparation nodes based on the encryption key, wherein any feedback information of the several feedback information sets includes: nickel content record data, doping element record data and nickel-rich layered transition metal oxide morphology record data; encode the nickel content record data and the doping element record data respectively through the trusted third party to obtain the nickel content coded record data and the doping element content coded record data; use the nickel-rich layered transition metal oxide morphology record data as supervision and the nickel content coded record data and the doping element content coded record data as input to train the nickel-rich layered transition metal oxide morphology prediction model, and deploy it on the trusted third party.
[0042] Specifically, a trusted third party refers to an independent and reliable organization or platform that is responsible for coordinating and managing tasks such as secure data transmission, encryption processing, and model training to ensure data privacy and security; ultra-high-performance lithium battery preparation nodes refer to various laboratories or production units involved in the lithium battery preparation process, such as manufacturers or research institutions, which can provide experimental data and receive instructions from a trusted third party; the feedback information set refers to the data set transmitted back from each preparation node, including information such as nickel content records, doping element records, and morphology records of nickel-rich layered transition metal oxides.
[0043] Specifically, the training process of the nickel-rich layered transition metal oxide morphology prediction model is as follows: first, a reliable third-party platform is established to coordinate tasks such as data collection, encryption, and model training; then, the doping element attribute set and encryption key are sent to multiple lithium battery preparation nodes through the trusted third party, and the multiple nodes encrypt the experimental data matching the doping element attribute set stored locally at the node according to the encryption key and then transmit it back to ensure the security of the data during transmission; then, the trusted third party receives the encrypted feedback information set, and performs the same encoding processing on the nickel content recording data and the doping element recording data therein as the aforementioned construction of the nickel content coding space and the doping element coding space, and converts these data into a format that can be processed by the model; then, the morphology recording data of the nickel-rich layered transition metal oxide is used as the supervision signal, and the encoded nickel content and doping element content data are used as input to train the morphology prediction model, and after the training is completed, the model is deployed on the trusted third-party platform for subsequent use.
[0044] The above steps achieve secure data collection and processing by introducing a trusted third party and data encryption mechanism, ensuring the data privacy and security of the participants. At the same time, the model training is carried out using data from multiple preparation nodes to improve the accuracy and generalization ability of the model, providing an efficient and reliable tool for subsequent composition optimization and morphology prediction.
[0045] In some implementations, the nickel-rich layered transition metal oxide morphology prediction model is trained using the nickel-rich layered transition metal oxide morphology record data as supervision and the nickel content encoded record data and the doping element content encoded record data as input, including:
[0046] The nickel-rich layered transition metal oxide morphology record data includes particle volume record data, particle shape record data, particle distribution morphology record data, particle surface roughness record data and particle surface active site distribution morphology record data; the particle volume record data is used as supervision, the nickel content coding record data and the doping element content coding record data are used as input to train the particle volume prediction unit; the particle shape record data is used as supervision, the nickel content coding record data and the doping element content coding record data are used as input to train the particle shape prediction unit; the particle distribution morphology record data is used as supervision, the nickel content coding record data and the doping element content coding record data are used as input to train the particle volume prediction unit. A particle distribution morphology prediction unit; using the particle surface roughness record data as supervision and the nickel content coding record data and the doping element content coding record data as input, training the particle surface roughness prediction unit; using the particle surface active site distribution morphology record data as supervision and the nickel content coding record data and the doping element content coding record data as input, training the particle surface active site distribution morphology prediction unit; merging the input layers of the particle volume prediction unit, the particle shape prediction unit, the particle distribution morphology prediction unit, the particle surface roughness prediction unit and the particle surface active site distribution morphology prediction unit to generate the nickel-rich layered transition metal oxide morphology prediction model.
[0047] Specifically, in the above steps, the nickel-rich layered transition metal oxide morphology recording data is used as a supervisory signal to guide the model to learn the mapping relationship between nickel content and doping element content and morphological characteristics.
[0048] Specifically, the nickel-rich layered transition metal oxide morphology prediction model includes multiple prediction units, where each prediction unit refers to an independent model module trained for a specific morphological feature (such as particle volume, particle shape, etc.), and each unit focuses on learning the relationship between a morphological feature and composition; input layer merging refers to combining the input layers of multiple prediction units to form a complete model architecture, so as to predict multiple morphological features at the same time.
[0049] Specifically, a prediction unit is trained for each morphological feature. For example, the particle volume prediction unit is trained with the particle volume recording data as the supervisory signal; the particle shape prediction unit is trained with the particle shape recording data as the supervisory signal; the particle distribution morphology prediction unit is trained with the particle distribution morphology recording data as the supervisory signal; and the particle surface roughness prediction unit is trained with the particle surface roughness recording data as the supervisory signal. Among them, the training inputs of multiple prediction units are the nickel content coded recording data and the doping element content coded recording data.
[0050] Furthermore, the input layers of all independent prediction units are merged to form a complete morphology prediction model. In this way, the model can process multiple morphological features at the same time and realize the prediction of the overall morphology of nickel-rich layered transition metal oxides.
[0051] The above method steps improve the model's prediction ability and accuracy for complex morphological features by decomposing the morphological features and training the prediction units separately; at the same time, the design of the merged input layer enables the model to comprehensively consider multiple morphological features, providing comprehensive prediction support for subsequent composition optimization.
[0052] Random encoding is performed according to the nickel content encoding space and the doping element encoding space to obtain component content encoding data, and the component content encoding data is processed by the nickel-rich layered transition metal oxide morphology prediction model to obtain the predicted morphology of the nickel-rich layered transition metal oxide.
[0053] Specifically, the nickel content coding space and the doping element coding space can be considered as the optimal selection space in the optimization process. By performing random coding in the nickel content coding space and the doping element coding space, a corresponding set of element content combinations with to-be-determined properties can be obtained. These combinations are then input into the morphology prediction model to predict the micromorphological characteristics of the corresponding nickel-rich layered transition metal oxide, including particle volume, shape, and distribution. This provides a basis for subsequent composition screening and optimization.
[0054] By performing random encoding in the coding space and using the prediction model to predict morphology, a large number of possible component combinations and their corresponding morphological features can be quickly generated, significantly improving the efficiency of component screening and reducing the consumption of time and resources; through the prediction model, the morphological features of the input component data can be quickly output, providing a scientific basis for component optimization, and further improving the efficiency of component configuration and the adaptability of batteries to application scenarios.
[0055] When the morphological similarity between the standard morphology of the nickel-rich layered transition metal oxide and the predicted morphology of the nickel-rich layered transition metal oxide is greater than or equal to a morphological similarity threshold, the component content encoding data is experimentally identified and returned to the client.
[0056] Specifically, the standard morphology is the ideal microstructural characteristics of nickel-rich layered transition metal oxides set by the user according to the application scenario requirements, including particle volume, shape, distribution pattern, surface roughness and active site distribution pattern; morphology similarity refers to the degree of matching between the predicted morphology and the standard morphology, which can be quantified by calculating the similarity indicators between the two (such as deviation, similarity coefficient, etc.); the morphology similarity threshold is a preset value used to judge whether the predicted morphology meets the user's required standards.
[0057] By setting a morphological similarity threshold and performing similarity judgment, we can quickly screen out component combinations that meet user needs, avoiding unnecessary experimental verification of a large number of components that do not meet the conditions, significantly improving the efficiency of component optimization, and returning the coded data of the component content that meets the conditions to the client, providing users with a clear experimental direction and reducing the blindness of the experiment.
[0058] In some embodiments, when the morphological similarity between the standard morphology of the nickel-rich layered transition metal oxide and the predicted morphology of the nickel-rich layered transition metal oxide is greater than or equal to a morphological similarity threshold, the component content coded data is experimentally marked and returned to the client, including:
[0059] The standard morphology of the nickel-rich layered transition metal oxide includes standard volume of particles, standard shape of particles, standard distribution morphology of particles, standard surface roughness of particles and standard distribution morphology of active sites; the predicted morphology of the nickel-rich layered transition metal oxide includes predicted volume of particles, predicted shape of particles, predicted distribution morphology of particles, predicted surface roughness of particles and predicted distribution morphology of active sites; when the volume deviation of the standard volume of the particles and the predicted volume of the particles is less than or equal to the volume deviation threshold, and the roughness deviation of the predicted surface roughness of the particles and the standard surface roughness of the particles is less than or equal to the roughness deviation threshold, and the particle shape similarity of the standard shape of the particles and the predicted shape of the particles is greater than or equal to the particle shape similarity threshold, and the particle distribution morphology similarity of the standard distribution morphology of the particles and the predicted distribution morphology of the particles is greater than or equal to the particle distribution morphology similarity threshold, and the active site distribution morphology similarity of the standard distribution morphology of the active sites and the predicted distribution morphology of the active sites is greater than or equal to the active site distribution morphology similarity threshold, the morphology similarity is deemed to be greater than or equal to the morphology similarity threshold; otherwise, the morphology similarity is deemed to be less than the morphology similarity threshold.
[0060] Specifically, the volume deviation threshold and roughness deviation threshold are preset numerical ranges used to determine whether the deviations in particle volume and surface roughness between the predicted and referenced morphologies are within acceptable limits. The particle shape similarity threshold, particle distribution morphology similarity threshold, and active site distribution morphology similarity threshold are used to determine whether the predicted and referenced morphologies meet the required similarity in shape, distribution, and active site distribution.
[0061] Specifically, through the twin neural network, the standard morphology of nickel-rich layered transition metal oxides and the predicted morphology of nickel-rich layered transition metal oxides are mapped to a feature space, and the distance or similarity between them is calculated to determine whether the two are similar. Among them, the twin neural network is trained with labeled sample comparison data, including positive samples marked as similar and negative samples marked as dissimilar. Through iterative training, the twin neural network can obtain the corresponding volume deviation threshold, roughness deviation threshold, particle shape similarity threshold, particle distribution morphology similarity threshold and active site distribution morphology similarity threshold.
[0062] The above-mentioned method steps introduce a twin neural network to calculate the similarity of particle shape, distribution morphology and active site distribution morphology, which can accurately evaluate the degree of match between the predicted morphology and the standard morphology, avoiding the misjudgment caused by inaccurate feature comparison in traditional methods; it can comprehensively and strictly screen out component combinations that meet user needs, further improving the accuracy and reliability of component optimization.
[0063] In some implementations, when the morphological similarity between the standard morphology of the nickel-rich layered transition metal oxide and the predicted morphology of the nickel-rich layered transition metal oxide is less than the morphological similarity threshold, a historical component content coded data set is obtained; the historical component content coded data set is sorted from large to small according to the morphological similarity to obtain a historical component content coded data sorting result; a first number of top-ranked coded data and a second number of bottom-ranked coded data of the historical component content coded data sorting result are obtained; the first number of top-ranked coded data is used as an update target to guide the second number of bottom-ranked coded data to mutate to obtain historical component content coded extended data; when the historical component content coded extended data meets the morphological similarity threshold, the historical component content coded extended data is experimentally marked and returned to the client.
[0064] Specifically, the historical component content coding data set includes all component content coding data previously generated in the optimization process and their corresponding morphological similarity records. The historical component content coding data set reflects the degree of matching between the predicted morphologies of different component combinations and the standard morphologies.
[0065] Specifically, when the predicted morphology does not meet the standard morphology requirements, first obtain the historical component content coding data set accumulated in the previous optimization process; then, sort the historical component content coding data set from large to small according to the morphology similarity recorded in the historical data; then, select a first number of coded data with higher rankings (i.e., data with higher morphology similarity) and a second number of coded data with lower rankings (i.e., data with lower morphology similarity) from the sorting results; further, use the coded data with higher rankings as the update target to guide the coded data with lower rankings to perform mutation operations to generate new component content coding expansion data, where mutation includes data exchange, modification, etc.
[0066] Furthermore, using the same method principle as mentioned above, the nickel-rich layered transition metal oxide morphology prediction model performs morphology prediction on the historical component content coded expansion data, and determines whether the obtained prediction results meet the morphology similarity threshold. If so, the corresponding historical component content coded expansion data is experimentally identified and returned to the client.
[0067] The above steps introduce historical component content coding data sets and merge them for sorting and mutation operations, so as to further explore component combinations that meet standard morphology requirements based on existing optimization experience. This not only makes full use of the information in historical data, but also introduces new possibilities through mutation operations, increasing the chance of finding better solutions, helping to speed up the process of component optimization, improve the adaptability of batteries and application scenarios, and at the same time reduce R&D costs and time.
[0068] In some implementations, the first number of coded data ranked higher than the first number is used as an update target to guide the second number of coded data ranked lower than the second number to mutate, thereby obtaining the historical component content coded expansion data, including:
[0069] Obtain first coded data of the first quantity sorted coded data, wherein the first coded data has a first coordinate in the nickel content coding space and a second coordinate in the doping element coding space; obtain second coded data of the second quantity sorted coded data, wherein the second coded data has a third coordinate in the nickel content coding space and a fourth coordinate in the doping element coding space; guide the variation of the third coordinate with the first coordinate as the target, and guide the variation of the fourth coordinate with the second coordinate as the target, to obtain first historical component content coded expansion data, and add it to the historical component content coded expansion data.
[0070] Specifically, the first coordinate and the second coordinate refer to the positions of the first quantity ranked coded data in the nickel content coding space and the doping element coding space, respectively, and these positions reflect the composition characteristics of high-similarity data; the third coordinate and the fourth coordinate refer to the positions of the second quantity ranked coded data in the nickel content coding space and the doping element coding space, respectively, and these positions reflect the composition characteristics of low-similarity data.
[0071] Specifically, the first coordinate and the second coordinate appear in pairs, and the number of pairs is consistent with the first number; the third coordinate and the fourth coordinate also appear in pairs, and the number of pairs is consistent with the second number; at the same time, the number of coordinate points in the second coordinate and the fourth coordinate is equal to the number of doping element types.
[0072] Specifically, with the first coordinate as the target, the third coordinate is mutated to bring it closer to the first coordinate. For example, if the first coordinate (nickel content) is 0.85 and the third coordinate is 0.80, the third coordinate can be adjusted to 0.83 through the mutation operation; with the second coordinate as the target, the fourth coordinate is mutated to bring it closer to the second coordinate. For example, if the second coordinate (doping element content) is 0.10 and the fourth coordinate is 0.08, the fourth coordinate can be adjusted to 0.09 through the mutation operation.
[0073] By using high-similarity data to guide low-similarity data to mutate, we can achieve targeted improvement of component combinations based on existing optimization experience, which can significantly improve optimization efficiency. At the same time, it provides more candidate solutions for subsequent model training and experimental verification, thereby improving the efficiency of component configuration and the adaptability of batteries to application scenarios.
[0074] In summary, the composition optimization method for an ultra-high performance lithium battery provided by the present invention has the following technical effects:
[0075] This system receives information about nickel content constraints, doping element constraints, and the standard morphology of nickel-rich layered transition metal oxides for ultra-high-performance lithium batteries from a client. Based on the nickel content constraints and doping element constraints, it constructs a nickel content encoding space and a doping element encoding space. Based on the characteristics of the doping element, it downloads a nickel-rich layered transition metal oxide morphology prediction model. This model is trained using multiple data sets, each containing nickel content encoding data, doping element content encoding data, and a label identifying the oxide morphology. Random encoding is performed within the nickel content encoding space and the doping element encoding space to obtain component content encoding data. This encoded data is then processed by the nickel-rich layered transition metal oxide morphology prediction model to obtain a predicted morphology for the nickel-rich layered transition metal oxide. When the similarity between the standard morphology of the nickel-rich layered transition metal oxide and the predicted morphology is greater than or equal to a preset morphology similarity threshold, the component content encoding data is identified as experimental data and returned to the client, thereby achieving the technical benefits of improving component configuration efficiency and optimizing battery and scenario compatibility.
[0076] Example 2, as Figure 2 , a structural diagram of a composition optimization system for an ultra-high performance lithium battery of the present invention. For example, Figure 1 The flow chart of the composition optimization method of an ultra-high performance lithium battery of the present invention can be shown as follows: Figure 2 The structure shown is implemented.
[0077] Based on the same concept as the composition optimization method of an ultra-high performance lithium battery in the above embodiment, the present invention also provides a composition optimization system for an ultra-high performance lithium battery, including:
[0078] The client data receiving component 11 is used to receive the nickel content constraint range, doping element constraint range and standard morphology of the nickel-rich layered transition metal oxide for ultra-high performance lithium batteries from the client.
[0079] The coding space construction component 12 is used to construct a nickel content coding space and a doping element coding space according to the nickel content constraint interval and the doping element constraint interval.
[0080] The morphology prediction model download component 13 is used to download a nickel-rich layered transition metal oxide morphology prediction model based on the doping element properties, wherein the nickel-rich layered transition metal oxide morphology prediction model is trained by multiple sets of data, and any set of the multiple sets of data includes: nickel content coded record data, doping element content coded record data and a label identifying the oxide morphology.
[0081] The predicted morphology acquisition component 14 is used to perform random encoding according to the nickel content encoding space and the doping element encoding space to obtain component content encoding data, and process the component content encoding data through the nickel-rich layered transition metal oxide morphology prediction model to obtain the predicted morphology of the nickel-rich layered transition metal oxide.
[0082] The experimental identification returning component 15 is used to return the component content encoding data to the client with experimental identification when the morphological similarity between the standard morphology of the nickel-rich layered transition metal oxide and the predicted morphology of the nickel-rich layered transition metal oxide is greater than or equal to the morphological similarity threshold.
[0083] In some embodiments, the code space construction component 12 includes:
[0084] The digit acquisition unit is used to obtain the maximum number of digits of the integer part and the maximum number of digits of the decimal part of the nickel content constraint interval.
[0085] A multidimensional coding space construction unit is used to construct a multidimensional coding space for the integer part according to the maximum number of digits of the integer part, and to construct a multidimensional coding space for the decimal part according to the maximum number of digits of the decimal part, wherein the dimension of the multidimensional coding space for the integer part is equal to the maximum number of digits of the integer part, and the dimension of the multidimensional coding space for the decimal part is equal to the maximum number of digits of the decimal part.
[0086] The nickel content integer part coding space acquisition unit is used to extract the integer part digit constraint interval according to the nickel content constraint interval, and perform regional restrictions on the integer part multidimensional coding space according to the integer part digit constraint interval to obtain the nickel content integer part coding space.
[0087] The nickel content fractional part coding space acquisition unit is used to extract the fractional part digit constraint interval according to the nickel content constraint interval, and perform regional restrictions on the fractional part multidimensional coding space according to the fractional part digit constraint interval to obtain the nickel content fractional part coding space.
[0088] The nickel content coding space setting unit is used to set the nickel content integer part coding space and the nickel content fractional part coding space as the nickel content coding space.
[0089] The construction process of the doping element code space is the same as that of the nickel content code space.
[0090] In some embodiments, the morphology prediction model downloading component 13 includes:
[0091] The trusted third party deployment unit is used to deploy a trusted third party.
[0092] A feedback information set receiving unit is used to send a doping element property set and an encryption key of a nickel-rich layered transition metal oxide to several ultra-high performance lithium battery preparation nodes through a trusted third party, and then receive several feedback information sets encrypted and returned by the several ultra-high performance lithium battery preparation nodes based on the encryption key, wherein any feedback information of the several feedback information sets includes: nickel content record data, doping element record data and nickel-rich layered transition metal oxide morphology record data.
[0093] The coded record data acquisition unit is used to encode the nickel content record data and the doping element record data respectively through a trusted third party to obtain the nickel content coded record data and the doping element content coded record data.
[0094] A morphology prediction model training and deployment unit is used to train the nickel-rich layered transition metal oxide morphology prediction model using the nickel-rich layered transition metal oxide morphology record data as supervision and the nickel content coded record data and the doping element content coded record data as input, and deploy it to the trusted third party.
[0095] In some implementations, the morphology prediction model training and deployment unit in the morphology prediction model download component 13 includes:
[0096] The morphology record data classification unit is used to clarify that the nickel-rich layered transition metal oxide morphology record data includes particle volume record data, particle shape record data, particle distribution morphology record data, particle surface roughness record data and particle surface active site distribution morphology record data.
[0097] The particle volume prediction unit training unit is used to train the particle volume prediction unit using the particle volume record data as supervision and the nickel content coded record data and the doping element content coded record data as input.
[0098] The particle shape prediction unit training unit is used to train the particle shape prediction unit using the particle shape record data as supervision and the nickel content coded record data and the doping element content coded record data as input.
[0099] The particle distribution morphology prediction unit training unit is used to train the particle distribution morphology prediction unit using the particle distribution morphology record data as supervision and the nickel content coded record data and the doping element content coded record data as input.
[0100] The particle surface roughness prediction unit training unit is used to train the particle surface roughness prediction unit using the particle surface roughness record data as supervision and the nickel content coded record data and the doping element content coded record data as input.
[0101] The particle surface active site distribution morphology prediction unit training unit is used to train the particle surface active site distribution morphology prediction unit using the particle surface active site distribution morphology record data as supervision and the nickel content coded record data and the doping element content coded record data as input.
[0102] A morphology prediction model generation unit is used to merge the input layers of the particle volume prediction unit, the particle shape prediction unit, the particle distribution morphology prediction unit, the particle surface roughness prediction unit and the particle surface active site distribution morphology prediction unit to generate the nickel-rich layered transition metal oxide morphology prediction model.
[0103] In some embodiments, the experiment identifier returning component 15 includes a shape similarity determination unit, configured to determine whether the shape similarity is greater than or equal to a shape similarity threshold according to the following conditions:
[0104] When the volume deviation between the standard volume of the particle and the predicted volume of the particle is less than or equal to the volume deviation threshold, and the roughness deviation between the predicted surface roughness of the particle and the standard surface roughness of the particle is less than or equal to the roughness deviation threshold, and the particle shape similarity between the standard shape of the particle and the predicted shape of the particle is greater than or equal to the particle shape similarity threshold, and the particle distribution morphology similarity between the standard distribution morphology of the particle and the predicted distribution morphology of the particle is greater than or equal to the particle distribution morphology similarity threshold, and the active site distribution morphology similarity between the standard distribution morphology of the active site and the predicted distribution morphology of the active site is greater than or equal to the active site distribution morphology similarity threshold, the morphology similarity is deemed to be greater than or equal to the morphology similarity threshold. Otherwise, the morphology similarity is deemed to be less than the morphology similarity threshold.
[0105] In some embodiments, the standard morphology of the nickel-rich layered transition metal oxide includes a standard volume of particles, a standard shape of particles, a standard distribution morphology of particles, a standard surface roughness of particles, and a standard distribution morphology of active sites. The predicted morphology of the nickel-rich layered transition metal oxide includes a predicted volume of particles, a predicted shape of particles, a predicted distribution morphology of particles, a predicted surface roughness of particles, and a predicted distribution morphology of active sites.
[0106] In some embodiments, the experiment identification returning component 15 further includes:
[0107] The historical component content coded data acquisition unit is used to obtain a historical component content coded data set when the morphological similarity between the standard morphology of the nickel-rich layered transition metal oxide and the predicted morphology of the nickel-rich layered transition metal oxide is less than the morphological similarity threshold.
[0108] The historical component content coded data sorting unit is used to sort the historical component content coded data set from large to small according to the morphological similarity to obtain the historical component content coded data sorting result.
[0109] The top and bottom sorting coding data acquisition unit is used to obtain a first number of top sorting coding data and a second number of bottom sorting coding data of the historical component content coding data sorting result.
[0110] The historical component content coded extended data generating unit is used to use the first quantity sorted earlier coded data as an update target, guide the second quantity sorted later coded data to mutate, and obtain the historical component content coded extended data.
[0111] The experiment identification returning unit is used to carry out an experiment identification on the historical component content encoded expanded data and return it to the client when the historical component content encoded expanded data meets the morphology similarity threshold.
[0112] In some implementations, the historical component content encoding expansion data generation unit in the experiment identifier return component 15 includes:
[0113] The first coded data acquisition unit is used to obtain first coded data of the first quantity sorted coded data, wherein the first coded data has a first coordinate in the nickel content coding space and a second coordinate in the doping element coding space.
[0114] The second coded data acquisition unit is used to obtain second coded data of the second quantity sorted later coded data, wherein the second coded data has a third coordinate in the nickel content coding space and a fourth coordinate in the doping element coding space.
[0115] The historical component content coded extended data generating unit is used to guide the third coordinate variation with the first coordinate as the target, and guide the fourth coordinate variation with the second coordinate as the target, to obtain the first historical component content coded extended data, and add it to the historical component content coded extended data.
[0116] It should be noted that, in the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant description of other embodiments.
[0117] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0118] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0119] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.
[0120] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.
[0121] Although preferred embodiments of the present invention have been described, additional changes and modifications to these embodiments may occur to those skilled in the art once the basic inventive concepts become known.
[0122] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the present invention and its equivalents, the present invention is also intended to include these modifications and variations.
Claims
1. A method for optimizing the composition of an ultra-high performance lithium battery, characterized in that: Applicable to servers, including: Receive from the client the nickel content constraint range, doping element constraint range, and standard morphology of nickel-rich layered transition metal oxides for ultra-high performance lithium batteries; Constructing a nickel content coding space and a doping element coding space according to the nickel content constraint interval and the doping element constraint interval; Downloading a nickel-rich layered transition metal oxide morphology prediction model based on doping element properties, wherein the nickel-rich layered transition metal oxide morphology prediction model is trained by multiple sets of data, and any set of the multiple sets of data includes: nickel content coded record data, doping element content coded record data, and a label identifying the oxide morphology; Random encoding is performed according to the nickel content encoding space and the doping element encoding space to obtain component content encoding data, and the component content encoding data is processed by the nickel-rich layered transition metal oxide morphology prediction model to obtain the predicted morphology of the nickel-rich layered transition metal oxide; When the morphological similarity between the standard morphology of the nickel-rich layered transition metal oxide and the predicted morphology of the nickel-rich layered transition metal oxide is greater than or equal to a morphological similarity threshold, the component content encoding data is experimentally identified and returned to the client.
2. The method according to claim 1, wherein Constructing a nickel content coding space and a doping element coding space according to the nickel content constraint interval and the doping element constraint interval includes: Obtaining the maximum number of digits of the integer part and the maximum number of digits of the decimal part of the nickel content constraint interval; Constructing a multidimensional encoding space for the integer part according to the maximum number of digits of the integer part, and constructing a multidimensional encoding space for the decimal part according to the maximum number of digits of the decimal part, wherein the dimension of the multidimensional encoding space for the integer part is equal to the maximum number of digits of the integer part, and the dimension of the multidimensional encoding space for the decimal part is equal to the maximum number of digits of the decimal part; Extracting the integer part digit constraint interval according to the nickel content constraint interval, and performing regional restrictions on the integer part multidimensional coding space according to the integer part digit constraint interval to obtain the nickel content integer part coding space; Extracting the decimal part constraint interval according to the nickel content constraint interval, and performing regional restriction on the decimal part multidimensional coding space according to the decimal part constraint interval to obtain the nickel content decimal part coding space; The nickel content integer part coding space and the nickel content fractional part coding space are set as the nickel content coding space; The construction process of the doping element code space is the same as that of the nickel content code space.
3. The method according to claim 1, wherein The nickel-rich layered transition metal oxide morphology prediction model training process includes: Deploy a trusted third party; After sending a set of doping element properties and an encryption key of a nickel-rich layered transition metal oxide to several ultra-high-performance lithium battery production nodes through a trusted third party, receiving several sets of feedback information encrypted and returned by the several ultra-high-performance lithium battery production nodes based on the encryption key, wherein any one of the several feedback information sets includes: nickel content record data, doping element record data, and nickel-rich layered transition metal oxide morphology record data; Encoding the nickel content record data and the doping element record data respectively through a trusted third party to obtain the nickel content coded record data and the doping element content coded record data; The nickel-rich layered transition metal oxide morphology record data is used as supervision, and the nickel content coded record data and the doping element content coded record data are used as input to train the nickel-rich layered transition metal oxide morphology prediction model and deploy it on the trusted third party.
4. The method according to claim 3, wherein The nickel-rich layered transition metal oxide morphology record data is used as supervision, and the nickel content coded record data and the doping element content coded record data are used as input to train the nickel-rich layered transition metal oxide morphology prediction model, including: The nickel-rich layered transition metal oxide morphology record data includes particle volume record data, particle shape record data, particle distribution morphology record data, particle surface roughness record data and particle surface active site distribution morphology record data; Using the particle volume record data as supervision and the nickel content encoded record data and the doping element content encoded record data as input, training a particle volume prediction unit; Using the particle shape record data as supervision and the nickel content encoded record data and the doping element content encoded record data as input, training a particle shape prediction unit; Using the particle distribution morphology recorded data as supervision and the nickel content coded recorded data and the doping element content coded recorded data as input, training a particle distribution morphology prediction unit; Using the particle surface roughness record data as supervision and the nickel content coded record data and the doping element content coded record data as input, training a particle surface roughness prediction unit; Using the particle surface active site distribution morphology record data as supervision and the nickel content coded record data and the doping element content coded record data as input, training a particle surface active site distribution morphology prediction unit; The input layers of the particle volume prediction unit, the particle shape prediction unit, the particle distribution morphology prediction unit, the particle surface roughness prediction unit and the particle surface active site distribution morphology prediction unit are merged to generate the nickel-rich layered transition metal oxide morphology prediction model.
5. The method according to claim 4, wherein When the morphological similarity between the standard morphology of the nickel-rich layered transition metal oxide and the predicted morphology of the nickel-rich layered transition metal oxide is greater than or equal to a morphological similarity threshold, the component content coded data is experimentally marked and returned to the client, including: The standard morphology of the nickel-rich layered transition metal oxide includes standard particle volume, standard particle shape, standard particle distribution morphology, standard particle surface roughness and standard active site distribution morphology; The predicted morphology of the nickel-rich layered transition metal oxide includes predicted particle volume, predicted particle shape, predicted particle distribution morphology, predicted particle surface roughness and predicted active site distribution morphology; When the volume deviation between the standard volume of the particle and the predicted volume of the particle is less than or equal to the volume deviation threshold, and the roughness deviation between the predicted surface roughness of the particle and the standard surface roughness of the particle is less than or equal to the roughness deviation threshold, and the particle shape similarity between the standard shape of the particle and the predicted shape of the particle is greater than or equal to the particle shape similarity threshold, and the particle distribution morphology similarity between the standard distribution morphology of the particle and the predicted distribution morphology of the particle is greater than or equal to the particle distribution morphology similarity threshold, and the active site distribution morphology similarity between the standard distribution morphology of the active site and the predicted distribution morphology of the active site is greater than or equal to the active site distribution morphology similarity threshold, the morphology similarity is deemed to be greater than or equal to the morphology similarity threshold; Otherwise, it is considered that the shape similarity is less than the shape similarity threshold.
6. The method according to claim 1, wherein Also includes: When the morphological similarity between the standard morphology of the nickel-rich layered transition metal oxide and the predicted morphology of the nickel-rich layered transition metal oxide is less than the morphological similarity threshold, obtaining a historical component content encoding data set; Sort the historical component content coded data set from largest to smallest according to the morphological similarity to obtain a historical component content coded data sorting result; Obtaining the first quantity of the encoded data ranked first and the second quantity of the encoded data ranked last in the sorting result of the historical component content encoded data; Taking the first number of coded data ranked first as the update target, guiding the second number of coded data ranked last to mutate, and obtaining historical component content coded expansion data; When the historical component content coded extended data meets the morphological similarity threshold, the historical component content coded extended data is subjected to experimental identification and returned to the client.
7. The method according to claim 6, wherein Taking the first number of coded data ranked higher as the update target, guiding the second number of coded data ranked lower to mutate, and obtaining historical component content coded expansion data, including: Obtaining first coded data of the first number of sorted coded data, wherein the first coded data has a first coordinate in the nickel content coding space and a second coordinate in the doping element coding space; Obtaining second coded data of the second number of coded data sorted later, wherein the second coded data has a third coordinate in the nickel content coding space and a fourth coordinate in the doping element coding space; The third coordinate is guided to vary with the first coordinate as a target, and the fourth coordinate is guided to vary with the second coordinate as a target, to obtain first historical component content coded extended data, and to add the first historical component content coded extended data.
8. A composition optimization system for ultra-high performance lithium batteries, characterized in that: The system is used to execute the composition optimization method of an ultra-high performance lithium battery according to any one of claims 1 to 7, and the system comprises: A client data receiving component is used to receive, from a client, nickel content constraint range, doping element constraint range, and standard morphology of nickel-rich layered transition metal oxides for ultra-high performance lithium batteries; A coding space construction component, configured to construct a nickel content coding space and a doping element coding space according to the nickel content constraint interval and the doping element constraint interval; A morphology prediction model download component is used to download a nickel-rich layered transition metal oxide morphology prediction model based on doping element properties, wherein the nickel-rich layered transition metal oxide morphology prediction model is trained by multiple sets of data, each of which includes: nickel content coded record data, doping element content coded record data, and a label identifying the oxide morphology; A predicted morphology acquisition component is used to perform random encoding according to the nickel content encoding space and the doping element encoding space to obtain component content encoding data, and process the component content encoding data using the nickel-rich layered transition metal oxide morphology prediction model to obtain the predicted morphology of the nickel-rich layered transition metal oxide; The experimental identification returning component is used to return the component content encoding data to the client with experimental identification when the morphological similarity between the standard morphology of the nickel-rich layered transition metal oxide and the predicted morphology of the nickel-rich layered transition metal oxide is greater than or equal to the morphological similarity threshold.