Solid hydrogen storage material design platform based on large language model and big data analysis

Abstract

The invention relates to the technical field of solid hydrogen storage material data analysis, in particular to a solid hydrogen storage material design platform based on a big language model and big data analysis. The analysis platform is realized based on HTML (Hypertext Markup Language), CSS (Cascading Style Sheet) and JavaScript, and comprises a homepage design module, a hydrogen storage material performance analysis intelligent atlas module, a hydrogen storage material data comparison visualization module, a solid hydrogen storage data recording and co-creating module and a hydrogen chat AI interactive question and answer module; and the homepage design module is used for constructing a solid hydrogen storage material data display and analysis page. According to the invention, a natural language, namely a query function, is developed based on a LangChain framework to help developers to construct an end-to-end application program by using a language model, and a set of tools, components and interfaces are also provided, so that the process of creating the application program supported by a large language model and a chat model can be simplified, and a full-dimensional intelligent hydrogen energy database architecture is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of solid-state hydrogen storage material data analysis, and in particular to a solid-state hydrogen storage material design platform based on large language models and big data analysis. Background Art

[0002] Amidst the accelerating global energy transition, solid-state hydrogen storage has emerged as a key hydrogen storage technology. Compared to traditional gaseous and liquid hydrogen storage, solid-state hydrogen storage has attracted significant attention due to its significant advantages. It achieves safe and efficient hydrogen storage by physically or chemically integrating hydrogen molecules into solid materials such as metal hydrides and organic frameworks. Its high volumetric hydrogen storage density, excellent safety, and good transport stability make it possible for large-scale hydrogen energy applications.

[0003] Through field research on hydrogen energy companies and scientific research institutions, it was found that the field of solid-state hydrogen storage has long faced the pain points of data dispersion, inefficient retrieval, and lack of analytical tools: manual integration of multi-source heterogeneous data and lack of intelligent analysis methods have seriously restricted technology research and development and results transformation. At present, solid-state hydrogen storage research continues to deepen, but its supporting data management platform is missing, and it is impossible to accurately obtain the type of materials required under specific conditions, resulting in researchers spending a lot of manpower and material resources in the early preparation stage to test the various properties of materials.

[0004] Based on this, the present invention provides a solid-state hydrogen storage material design platform based on large language models and big data analysis to solve the technical problems raised above. Summary of the Invention

[0005] The purpose of the present invention is to provide a solid-state hydrogen storage material design platform based on large language models and big data analysis to solve the problems raised by the above background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A solid-state hydrogen storage material design platform based on large language models and big data analysis is provided. The analysis platform is implemented in HTML, CSS, and JavaScript, and includes a homepage design module, an intelligent graph module for hydrogen storage material performance analysis, a hydrogen storage material data comparison and visualization module, a solid-state hydrogen storage data collection and co-creation module, and a "Hydrogen Chat" AI interactive question-and-answer module.

[0008] The homepage design module is used to construct a data display and analysis page for solid-state hydrogen storage materials;

[0009] The hydrogen storage material performance analysis intelligent graph module is used to draw and display hydrogen storage material performance analysis graphs through jQuery and ECharts data visualization web scripts;

[0010] The hydrogen storage material data comparison and visualization module uses data visualization technology to conduct in-depth comparative analysis of existing hydrogen storage materials and previous research data;

[0011] The solid-state hydrogen storage data collection and co-creation module is used for continuous updating and optimization of the hydrogen storage material database through manual input and machine learning technology.

[0012] The "Hydrogen Chat" AI interactive question-and-answer module uses the LangChain framework, pre-trained models, Pandas, and Plotly.express to intelligently analyze data and generate visual reports in the hydrogen energy field.

[0013] Preferably, the homepage design module further includes a navigation bar, a data module, a material atlas display module, a dynamic data visualization module and a material update dynamic module;

[0014] The navigation bar is used to provide multi-language switching and login;

[0015] The data module is used for data set introduction and data volume statistics;

[0016] The material atlas display module is used to display the performance comparison of different materials in a scatter plot;

[0017] The dynamic data visualization module is used to generate pie charts of the number of papers and the proportion of material categories;

[0018] The material update dynamic module is used to update material data and characteristics.

[0019] Preferably, the hydrogen storage material performance analysis intelligent graph module further includes a design area and a graph output area;

[0020] The design area is used to provide multiple parameter coordinate axes, realize one-click drawing and annotation, and annotate the dehydrogenation peak temperature after the drawing is completed;

[0021] The drawing output area is used to support personalized adjustment of graph line colors and data point display categories, conduct horizontal comparison of material performance parameters, and support the download of high-quality scientific research drawings.

[0022] Preferably, in the hydrogen storage material performance analysis intelligent atlas module, the implementation steps of intelligently marking its dehydrogenation peak temperature are:

[0023] S1. First, enter the experimental data into the platform, including the key parameters of time on the X-axis and hydrogen content on the Y-axis;

[0024] S2. Automatically draw the dehydrogenation performance curve based on jQuery and ECharts data visualization web script;

[0025] S3. During the mapping process, an algorithm automatically detects the peak point on the curve, i.e., the dehydrogenation peak temperature;

[0026] S4. Once a peak is detected, the location of the peak point and its corresponding dehydrogenation peak temperature are automatically marked on the chart, allowing researchers to intuitively understand the dehydrogenation performance of the material.

[0027] Preferably, the hydrogen storage material data comparison and visualization module allows users to customize the data screening range, provides dynamic interactive functions, highlights the key characteristics of specific materials, and supports personalized data annotation to help researchers quickly identify and analyze the advantages or disadvantages of material performance.

[0028] Preferably, the steps for constructing the solid-state hydrogen storage data collection and co-creation module are:

[0029] S11. Support users to upload experimental data and literature data;

[0030] S12. Automatically extract key parameters and generate a structured database through machine learning, clean and normalize heterogeneous data, and establish a unified material characterization system.

[0031] Preferably, the implementation steps of developing a multi-dimensional deep learning framework in step S13 are:

[0032] S1311. Develop a deep learning framework specifically for automatic data entry of hydrogen energy materials by integrating multi-dimensional information from microscopic images and spectral data of materials.

[0033] S1312. Use a large amount of labeled hydrogen energy material data to train a deep learning framework to accurately identify and extract key characteristic parameters of the materials;

[0034] S1313. Apply the trained deep learning framework to the actual data entry process to achieve automated and intelligent data processing.

[0035] Preferably, the step of connecting upstream and downstream data in step S13 is:

[0036] S1321. Develop a dedicated data interface for interfacing with data systems in the upstream and downstream industries of water electrolysis and fuel cell production;

[0037] S1322. Establish unified data mapping rules based on the characteristics and needs of upstream and downstream data to ensure the accuracy and consistency of data during transmission;

[0038] S1323. Integrate and analyze the connected upstream and downstream data with the data in the hydrogen energy material big data intelligent analysis platform to build an analysis map of the entire hydrogen energy industry chain.

[0039] Preferably, the implementation process of the "Hydrogen Chat" AI interactive question-and-answer module is as follows:

[0040] S21. Develop a "natural language query" function based on the LangChain framework to enable users to generate standard SQL statements and extract data from their daily language questions.

[0041] S22. Simultaneously, an intelligent visual analysis engine is built to enable real-time processing of query results and generation of visual charts, which in turn generates structured reports, automating the entire "questioning - analysis - charting - reporting" process.

[0042] S23. Innovative design of directional channel direct connection visualization module and "data sandbox" security architecture to ensure data security and analysis capabilities.

[0043] Preferably, the implementation steps of the data sandbox in step S23 are:

[0044] S231. The user question and database structure information are combined and used only to generate SQL commands, while the actual data does not pass through the external large model or is protected from unauthorized access;

[0045] S232. All data analysis and visualization operations are performed locally to ensure that data is not leaked to the external environment;

[0046] S233. Adopt security policies and access control mechanisms to ensure that only authorized users can access and process data in the data sandbox.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] This invention develops a natural language query function based on the LangChain framework to help developers build end-to-end applications using language models. It also provides a set of tools, components, and interfaces that simplify the process of creating applications powered by large language models and chat models. The full-dimensional intelligent hydrogen energy database architecture and the hydrogen energy material big data intelligent analysis platform are a multi-source data fusion system. By integrating data from key areas such as solid-state hydrogen storage and electrochemistry, the platform covers a precise data pool from multiple papers, supports the mining of material performance patterns and intelligent screening, and features dynamic updates and cross-analysis. The platform updates data in real time and supports the linkage analysis of multi-dimensional parameters. Users can quickly locate high-potential materials through dynamic visualization tools such as scatter plots and bar charts. The "Hydrogen Chat" platform integrates the entire process of data analysis, visualization, and report generation using a large language model, embeds hydrogen energy domain knowledge models, and provides performance trend predictions and intelligent material recommendations. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1This is a block diagram of the overall system of the present invention;

[0050] Figure 2 Draw a flow chart for the intelligent map of the present invention;

[0051] Figure 3 This is an example diagram of data comparison of the present invention;

[0052] Figure 4 This is the interface diagram of the co-creation and collection module of the present invention;

[0053] Figure 5 An example diagram of a web page built for the front end of the present invention;

[0054] Figure 6 This is the PCT image of the present invention;

[0055] Figure 7 This is the webplot software interface of the present invention. DETAILED DESCRIPTION

[0056] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the embodiments described 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 ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0057] Example 1, please refer to Figures 1 to 7 The present invention proposes a solid-state hydrogen storage material design platform based on large language models and big data analysis. Furthermore, the analysis platform is implemented based on HTML, CSS, and JavaScript, including a homepage design module, an intelligent graph module for hydrogen storage material performance analysis, a hydrogen storage material data comparison and visualization module, a solid-state hydrogen storage data collection and co-creation module, and a "Hydrogen Chat" AI interactive question-and-answer module.

[0058] Among them, it should be noted that the homepage design module is used to construct the data display and analysis page of solid-state hydrogen storage materials. The hydrogen storage material performance analysis intelligent map module uses jQuery and ECharts data visualization web scripts to draw and display hydrogen storage material performance analysis maps. The hydrogen storage material data comparison and visualization module uses data visualization technology for in-depth comparative analysis of existing hydrogen storage materials and previous research data. The solid-state hydrogen storage data collection and co-creation module uses manual entry and machine learning technology to continuously update and optimize the hydrogen storage material database. The "Hydrogen Chat" AI interactive question and answer module uses the LangChain framework, pre-trained models, Pandas and Plotly.express for intelligent analysis of hydrogen energy data and generation of visual reports.

[0059] In this embodiment, it should also be noted that the homepage design module also includes a navigation bar, a data module, a material map display module, a dynamic data visualization module, and a material update dynamic module;

[0060] Furthermore, the navigation bar is used to provide multi-language switching and login;

[0061] Furthermore, the data module is used for data set introduction and data volume statistics;

[0062] Furthermore, the material map display module is used to display the performance comparison of different materials in scatter plots;

[0063] Furthermore, the dynamic data visualization module is used to generate pie charts showing the trend of the number of papers and the proportion of material categories;

[0064] Furthermore, the material update dynamic module is used to update material data and characteristics;

[0065] In this embodiment, it should also be noted that the hydrogen storage material performance analysis intelligent map module also includes a design area and a map output area;

[0066] Furthermore, the design area is used to provide multiple parameter coordinate axes, enabling one-click drawing and annotation, and annotating the dehydrogenation peak temperature after the drawing is completed;

[0067] Furthermore, the output area supports personalized adjustment of graph line colors and data point display categories, horizontal comparison of material performance parameters, and supports the download of high-quality scientific research drawings;

[0068] In this embodiment, it should also be noted that in the hydrogen storage material performance analysis intelligent atlas module, the implementation steps of intelligently marking its dehydrogenation peak temperature are as follows:

[0069] S1. First, enter the experimental data into the platform, including the key parameters of time on the X-axis and hydrogen content on the Y-axis;

[0070] S2. Automatically draw the dehydrogenation performance curve based on jQuery and ECharts data visualization web script;

[0071] S3. During the mapping process, an algorithm automatically detects the peak point on the curve, i.e., the dehydrogenation peak temperature;

[0072] S4. Once a peak is detected, the peak location and its corresponding dehydrogenation peak temperature are automatically marked on the graph, allowing researchers to intuitively understand the material's dehydrogenation performance.

[0073] In this embodiment, it should also be noted that the hydrogen storage material data comparison and visualization module allows users to customize the data screening range, provides dynamic interactive functions, highlights the key characteristics of specific materials, and supports personalized data annotation, helping researchers to quickly identify and analyze the advantages or disadvantages of material performance;

[0074] In this embodiment, it should also be noted that the steps for constructing the solid-state hydrogen storage data collection and co-creation module are as follows:

[0075] S11. Support users to upload experimental data and literature data;

[0076] S12. Automatically extract key parameters and generate a structured database through machine learning, clean and normalize heterogeneous data, and establish a unified material characterization system;

[0077] In this embodiment, it should also be noted that the implementation steps of developing a multi-dimensional deep learning framework in step S13 are:

[0078] S1311. Develop a deep learning framework specifically for automatic data entry of hydrogen energy materials by integrating multi-dimensional information from microscopic images and spectral data of materials.

[0079] S1312. Use a large amount of labeled hydrogen energy material data to train a deep learning framework to accurately identify and extract key characteristic parameters of the materials;

[0080] S1313. Apply the trained deep learning framework to the actual data entry process to achieve automated and intelligent data processing;

[0081] In this embodiment, it should also be noted that the steps of connecting upstream and downstream data in step S13 are:

[0082] S1321. Develop a dedicated data interface for interfacing with data systems in the upstream and downstream industries of water electrolysis and fuel cell production;

[0083] S1322. Establish unified data mapping rules based on the characteristics and needs of upstream and downstream data to ensure the accuracy and consistency of data during transmission;

[0084] S1323. Integrate and analyze the connected upstream and downstream data with the data in the hydrogen energy material big data intelligent analysis platform to build an analysis map of the entire hydrogen energy industry chain;

[0085] In this embodiment, it should also be noted that the implementation process of the "Hydrogen Chat" AI interactive question-and-answer module is as follows:

[0086] S21. Develop a "natural language query" function based on the LangChain framework to enable users to generate standard SQL statements and extract data from their daily language questions.

[0087] S22. Simultaneously, an intelligent visual analysis engine is built to enable real-time processing of query results and generation of visual charts, which in turn generates structured reports, automating the entire "questioning - analysis - charting - reporting" process.

[0088] S23. Innovative design of directional channel direct connection visualization module and "data sandbox" security architecture to ensure data security and analytical capabilities;

[0089] In this embodiment, it should be noted that the implementation steps of the data sandbox in step S23 are:

[0090] S231. The user question and database structure information are combined and used only to generate SQL commands, while the actual data does not pass through the external large model or is protected from unauthorized access;

[0091] S232. All data analysis and visualization operations are performed locally to ensure that data is not leaked to the external environment;

[0092] S233. Use security policies and access control mechanisms to ensure that only authorized users can access and process data in the data sandbox;

[0093] Example 2: In actual application, the application of the artificial intelligence-based hydrogen energy material big data intelligent analysis platform includes the following process:

[0094] See also Figure 1 As shown, the present invention provides a solid-state hydrogen storage material design platform based on large language models and big data analysis, including a homepage design module, a hydrogen storage material performance analysis intelligent map module, a hydrogen storage material data comparison and visualization module, a solid-state hydrogen storage data collection module, and a "hydrogen chat" AI interactive question and answer module;

[0095] In one embodiment of the present invention, the homepage design module includes a navigation bar, a data module, a material map display, dynamic data visualization, and a material update dynamic section, which is intended to provide data display and analysis of solid-state hydrogen storage materials;

[0096] Specifically, it provides multi-language switching (Chinese and English) and login functions, supports quick module switching, displays the current page in bright colors, supports switching between Chinese and English, and facilitates use by users of different languages, improving the global user experience;

[0097] Through the login function, ensure user data security, prevent unauthorized access, analyze user behavior, and optimize product functions;

[0098] The data module consists of two parts: data set introduction and data volume statistics. By combining the data set introduction and data volume statistics, users can fully understand the background, scale and quality of the data, and enhance the credibility and transparency of the data.

[0099] The material map display section uses scatter plots to compare the performance of different materials. It can clearly show the relationship between two performance indicators and help users quickly understand the distribution trend of data. In addition, it can intuitively and efficiently reveal data distribution, patterns and trends, supporting multi-dimensional analysis and data-driven decision-making.

[0100] The dynamic data visualization section displays the number of materials-related papers published in recent years, updates data in real time, and intuitively demonstrates research hotspots and development trends in the field of materials science;

[0101] By counting the proportion of current material categories in the database and presenting it visually through pie charts, we can quickly identify the dominant material categories in the database, help understand the proportion of each type of material, and help to rationally allocate research resources and give priority to supporting major categories or potential categories;

[0102] The Materials Update section provides the latest updated material data and properties, providing a reliable basis for material selection and process optimization decisions, ensuring the accuracy and reliability of data, improving the scientificity and accuracy of decision-making, helping researchers grasp the latest research progress, more quickly identify new materials or new properties, and accelerate the innovation process;

[0103] In one embodiment of the present invention, the hydrogen storage material performance analysis intelligent graph module includes a design area and a graph output area, aiming to provide a smooth user experience, efficient data presentation, and good visual effects, and can greatly simplify the data processing workflow of researchers;

[0104] Specifically, in the drawing design area, the platform provides a variety of different parameter coordinate axes. Users can select appropriate coordinate axes to simplify the analysis of complex data, quickly identify trends and anomalies, and enhance the flexibility and functionality of the platform;

[0105] Take the dehydrogenation performance curve of a magnesium-based material as an example, see Figure 2 As shown, the X-axis is time and the Y-axis is hydrogen content. After entering the experimental data, you can draw a graph with one click. After the graph is drawn, its dehydrogenation peak temperature will be intelligently marked. In addition, the graph line color (to distinguish different experimental conditions) and data point display category can be adjusted according to the personalized needs of researchers.

[0106] By calibrating different screening methods in the upper and lower halves of the graph area, different performance parameters can be extracted for horizontal comparison. The advantages and disadvantages of the dehydrogenation platform pressure and peak temperature performance parameters after modification compared with various materials can be tested. By using different screening methods, key performance parameters can be quickly extracted, reducing data processing time and improving analysis efficiency. Comparison under different screening methods can more clearly display data characteristics, enhance visualization effects, and make it easier to discover potential problems or anomalies.

[0107] At the same time, the performance parameters under different screening conditions are compared horizontally to intuitively display the differences and facilitate the identification of advantages and disadvantages;

[0108] Finally, users can directly obtain high-quality scientific research drawings by downloading images to their local computer, which greatly simplifies the data processing workflow of researchers and greatly improves efficiency.

[0109] See also Figure 3 As shown, in one embodiment of the present invention, the hydrogen storage material data comparison and visualization module is designed to use data visualization technology to conduct in-depth comparative analysis of existing hydrogen storage materials and previous research data, and flexibly adjust the data presentation method based on the specific needs of users to provide clear and intuitive material performance evaluation;

[0110] Specifically, data visualization technology is used to conduct in-depth comparative analysis of existing hydrogen storage materials and previous research data. Data visualization can convert complex data into intuitive charts and graphs. Users can intuitively see the key indicators of hydrogen storage capacity and release rate of different hydrogen storage materials, as well as how these indicators change with temperature and pressure conditions. This helps researchers quickly identify hydrogen storage materials with excellent performance and provides strong support for subsequent research and development. In combination with the specific needs of users, the data presentation method can be flexibly adjusted to provide clear and intuitive material performance evaluation.

[0111] Through various visualization methods such as line graphs and bar charts, this module can show the relative position of a single material in the entire hydrogen storage material system. Researchers can effectively compare existing hydrogen storage materials with previous research data. This comparison is not limited to a single indicator, but can also be multi-dimensional and comprehensive. For example, researchers can use visual charts to simultaneously display the hydrogen storage capacity, release rate, and stability of multiple hydrogen storage materials, thereby comprehensively evaluating the comprehensive performance of different materials.

[0112] The module allows users to customize data filtering ranges and provides dynamic interactive functions to enhance data readability and comparison, enabling users to adjust visualization methods for different research needs and highlight the key characteristics of specific materials;

[0113] In addition, the system supports personalized data annotation, which can add unique identifiers to specific data points in the chart and provide special instructions in the legend, helping researchers quickly identify and analyze the performance advantages or disadvantages of a material and make more accurate horizontal comparisons with other materials.

[0114] See also Figure 4 As shown, in one embodiment of the present invention, the solid-state hydrogen storage data collection and co-creation module is intended to continuously update and optimize the hydrogen storage material database through user participation and AI technology to support the development of the entire hydrogen energy industry chain;

[0115] Specifically, data collection supports users to upload experimental data (such as crystal structure, adsorption kinetics curve) and literature data through manual entry;

[0116] After the database matures, the data is standardized through machine learning. Standardization can convert data of different dimensions to the same scale, making different features or variables comparable and consistent. This helps to more accurately evaluate the various characteristics of the material in subsequent data analysis and model training.

[0117] Compatible with Excel, CSV, and PDF formats, it automatically extracts key parameters and generates a structured database. Compatible with files in various formats, it allows data to be easily exchanged and shared between different applications and systems.

[0118] Through automated extraction, key parameters can be quickly extracted from files of various formats, avoiding the tedious and time-consuming manual input;

[0119] The extracted key parameters can be directly imported into the structured database, eliminating the need for additional data sorting, thus improving the efficiency of database generation;

[0120] AI algorithms are used to clean and normalize heterogeneous data and establish a unified material characterization system. AI algorithms use natural language processing and image recognition technology to automatically align and integrate data, reducing the cost of manual intervention and thus R&D costs.

[0121] The unified material characterization system provides more convenient and efficient data support for material research and development, accelerating the screening, optimization and verification of new materials;

[0122] In one embodiment of the present invention, the "Hydrogen Chat" AI interactive question-and-answer module aims to achieve intelligent analysis of hydrogen energy data through natural language processing technology. Its core technical route is built around two core capabilities;

[0123] Specifically, the "natural language query" function is developed based on the LangChain framework. When users ask questions in everyday language, the system generates SQL statements that comply with database specifications through structured prompt templates and directly extracts data from the hydrogen energy knowledge base.

[0124] During this process, we optimized our understanding of industry terminology through pre-training models and ensured the accuracy of generated SQL syntax through a context-sensitive dynamic constraint framework, solving the problem of complex database queries.

[0125] The SQL statements generated by the structured prompt templates comply with database specifications and therefore have good cross-platform compatibility, meaning that the same SQL statements can be executed on different database management systems, thereby improving data accessibility.

[0126] The data extracted directly from the hydrogen energy knowledge base ensures data consistency and integrity, avoiding data inconsistency issues between different sources;

[0127] While implementing the general question-and-answer module, an intelligent visual analysis engine is built. Query results are directly transmitted to the central data structure of the Pandas database open source library through LangChain for real-time processing. Plotly.express then automatically generates visual charts and structured reports, automating the entire "question-analysis-chart-report" process.

[0128] Through the intelligent question-and-answer module, users can quickly submit queries without having to manually write complex query statements or data analysis code. LangChain can automatically parse the query and transfer the results to the central data structure of the Pandas database open source library, eliminating the tedious steps of data import and preprocessing. The central data structure of the Pandas database open source library provides rich data processing and analysis functions, which can quickly complete data cleaning, conversion, and aggregation operations.

[0129] In addition, Plotly.express supports a variety of chart types, such as line charts, scatter charts, and bar charts, which can meet different data visualization needs, optimize user experience, and promote data-driven decision-making;

[0130] The following describes in detail the construction method and key technologies of the artificial intelligence-based hydrogen energy material big data intelligent analysis platform of the present application by way of example. It should be understood that this example does not constitute any limitation to the present application;

[0131] See also Figure 5 As shown, in this embodiment, the construction of the front-end web page relies on Hypertext Markup Language, Cascading Style Sheets and Java scripts;

[0132] In this example, the project uses VS Code, a cross-platform source code editor, and Vite, a front-end to build a network to quickly initialize the project structure.

[0133] After the project is created, it mainly includes Hypertext Markup Language structure, Cascading Style Sheets and Java script interaction logic;

[0134] Hypertext Markup Language is responsible for organizing page content, Cascading Style Sheets are used to beautify the interface, and Java scripts control dynamic interactions;

[0135] Take the catalysis, alloying and doping modules as an example;

[0136] In this embodiment, the DataCollection.html page collects the data submitted by the user by calling the connector.js script and sends it to the backend interface (https: / / s-nydrogendataplatform.nas.npolarcn / api / add new catalyst);

[0137] After receiving the data, the backend will first verify whether the document’s DOI already exists in the database;

[0138] If the DOI is not found in the database, the backend will return a message of failed entry to the frontend;

[0139] If the DOI already exists, the system will create a CatalyticAndFormula class and store the relevant data in the class in the database after mapping it using object-relational mapping technology;

[0140] At the same time, the backend will return a successful entry message to the frontend;

[0141] In this embodiment, the backend of the project uses object-relational mapping technology to avoid directly using the MySQL dialect, thereby effectively preventing SQL injection attacks and reducing the risk of malicious operations on the database;

[0142] Using object-relational mapping technology not only enhances data security, but also improves code maintainability and scalability;

[0143] In this embodiment, the backend also uses the Python Flask framework, which enables the database to maintain good performance in the case of high concurrent reading and writing, ensuring stability and response speed when multiple people operate simultaneously;

[0144] The combination of the two improves system performance while ensuring data security and consistency.

[0145] In this embodiment, the data extraction part includes extracting thermodynamic data and kinetic data of the material. The thermodynamic data includes the PCT curve (pressure-composition-temperature curve), DSC curve (differential scanning calorimetry) and TPD (programmed temperature curve) of the material, and the kinetic data is mainly the hydrogen absorption and desorption time curve of the material.

[0146] Taking the PCT curve as an example, data extraction can be roughly divided into four steps: First, we need to find the PCT curve of relevant materials in the literature, see Figure 6 As shown, the curve is saved as a picture by taking a screenshot;

[0147] Secondly, use origin and webplot software to extract data points. Open the saved curve image in origin and webplot software, and get the horizontal and vertical coordinates of the data points on the curve by setting the horizontal and vertical axes. Take webplot as an example, see Figure 7 As shown;

[0148] Then check the data obtained from the points you took and use Excel software to calculate the average value of the ordinate (pressure);

[0149] Finally, the data is entered into the hydrogen energy material big data intelligent analysis platform through the database co-creation module;

[0150] The biggest advantage of this solution is its full-dimensional intelligent hydrogen energy database architecture. The hydrogen energy material big data intelligent analysis platform is a multi-source data fusion system. By integrating data from key areas such as solid-state hydrogen storage and electrochemistry (HER / OER / ORR), the platform covers a precise data pool of 1,826 papers, supporting the mining of material performance patterns and intelligent screening.

[0151] The platform has the characteristics of dynamic update and cross-analysis;

[0152] The platform updates data in real time (e.g., 522 new Mg-based data items were added in 2024) and supports linkage analysis of multi-dimensional parameters (hydrogen storage capacity, dehydrogenation temperature);

[0153] Users can quickly locate high-potential materials through dynamic visualization tools such as scatter plots and histograms;

[0154] AI-driven natural language interaction, the "Hydrogen Chat" platform integrates a large language model for the entire process of data analysis, visualization and report generation, and embeds knowledge models in the hydrogen energy field to provide performance trend predictions and intelligent material recommendations.

[0155] Throughout this specification, references to "one embodiment," "example," or "specific example" indicate that a specific feature, structure, material, or characteristic described in conjunction with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of these terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0156] The preferred embodiments of the present invention disclosed above are intended only to help illustrate the present invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the present invention to the specific embodiments described. Obviously, many modifications and variations are possible based on the content of this specification. These embodiments are selected and described in detail in this specification to better explain the principles and practical applications of the present invention, thereby enabling those skilled in the art to better understand and utilize the present invention. The present invention is limited only by the claims and their full scope and effectiveness.

Claims

1. A solid-state hydrogen storage material design platform based on large language models and big data analysis, the analysis platform is implemented based on HTML, CSS, and JavaScript, and is characterized by: It includes a homepage design module, an intelligent graph module for hydrogen storage material performance analysis, a hydrogen storage material data comparison and visualization module, a solid-state hydrogen storage data collection and co-creation module, and a "Hydrogen Chat" AI interactive question-and-answer module; The homepage design module is used to construct a data display and analysis page for solid-state hydrogen storage materials; The hydrogen storage material performance analysis intelligent graph module is used to draw and display hydrogen storage material performance analysis graphs through jQuery and ECharts data visualization web scripts; The hydrogen storage material data comparison and visualization module uses data visualization technology to conduct in-depth comparative analysis of existing hydrogen storage materials and previous research data; The solid-state hydrogen storage data collection and co-creation module uses manual input and machine learning technology to continuously update and optimize the hydrogen storage material database. The "Hydrogen Chat" AI interactive question-and-answer module uses the LangChain framework, pre-trained models, Pandas, and Plotly.express to intelligently analyze data and generate visual reports in the hydrogen energy field.

2. The solid-state hydrogen storage material design platform based on large language model and big data analysis according to claim 1 is characterized in that: The homepage design module also includes a navigation bar, a data module, a material map display module, a dynamic data visualization module and a material update dynamic module; The navigation bar is used to provide multi-language switching and login; The data module is used for data set introduction and data volume statistics; The material atlas display module is used to display the performance comparison of different materials in a scatter plot; The dynamic data visualization module is used to generate pie charts of the number of papers and the proportion of material categories; The material update dynamic module is used to update material data and characteristics.

3. The solid-state hydrogen storage material design platform based on large language model and big data analysis according to claim 1 is characterized in that: The hydrogen storage material performance analysis intelligent graph module also includes a design area and a graph output area; The design area is used to provide multiple parameter coordinate axes, realize one-click drawing and annotation, and annotate the dehydrogenation peak temperature after the drawing is completed; The drawing output area is used to support personalized adjustment of graph line colors and data point display categories, conduct horizontal comparison of material performance parameters, and support the download of high-quality scientific research drawings.

4. The solid-state hydrogen storage material design platform based on large language model and big data analysis according to claim 1 is characterized in that: In the hydrogen storage material performance analysis intelligent atlas module, the implementation steps for intelligently marking its dehydrogenation peak temperature are as follows: S1. First, enter the experimental data into the platform, including the key parameters of time on the X-axis and hydrogen content on the Y-axis; S2. Automatically draw the dehydrogenation performance curve based on jQuery and ECharts data visualization web script; S3. During the mapping process, an algorithm is used to automatically detect the peak point on the curve, i.e., the dehydrogenation peak temperature; S4. Once a peak is detected, the location of the peak point and its corresponding dehydrogenation peak temperature are automatically marked on the chart, allowing researchers to intuitively understand the dehydrogenation performance of the material.

5. The solid-state hydrogen storage material design platform based on large language model and big data analysis according to claim 1 is characterized in that: The hydrogen storage material data comparison and visualization module allows users to customize data screening ranges, provides dynamic interactive functions, highlights the key characteristics of specific materials, and supports personalized data annotation, helping researchers quickly identify and analyze material performance advantages or disadvantages.

6. The solid-state hydrogen storage material design platform based on large language model and big data analysis according to claim 1 is characterized in that: The steps for constructing the solid-state hydrogen storage data collection and co-creation module are as follows: S11. Support users to upload experimental data and literature data; S12. Automatically extract key parameters and generate a structured database through machine learning, clean and normalize heterogeneous data, and establish a unified material characterization system.

7. The solid-state hydrogen storage material design platform based on large language model and big data analysis according to claim 6 is characterized in that: The implementation steps of developing a multi-dimensional deep learning framework in step S13 are: S1311. Develop a deep learning framework specifically for automatic data entry of hydrogen energy materials by integrating multi-dimensional information from microscopic images and spectral data of materials. S1312. Use a large amount of labeled hydrogen energy material data to train a deep learning framework to accurately identify and extract key characteristic parameters of the materials; S1313. Apply the trained deep learning framework to the actual data entry process to achieve automated and intelligent data processing.

8. The solid-state hydrogen storage material design platform based on large language model and big data analysis according to claim 6 is characterized in that: The steps of connecting upstream and downstream data in step S13 are: S1321. Develop a dedicated data interface for interfacing with data systems in the upstream and downstream industries of water electrolysis and fuel cell production; S1322. Establish unified data mapping rules based on the characteristics and needs of upstream and downstream data to ensure the accuracy and consistency of data during transmission; S1323. Integrate and analyze the connected upstream and downstream data with the data in the hydrogen energy material big data intelligent analysis platform to build an analysis map of the entire hydrogen energy industry chain.

9. The solid-state hydrogen storage material design platform based on large language model and big data analysis according to claim 8, characterized in that: The implementation process of the "Hydrogen Chat" AI interactive question-and-answer module is as follows: S21. Develop a "natural language query" function based on the LangChain framework to enable users to generate standard SQL statements and extract data from their daily language questions. S22. Simultaneously, an intelligent visual analysis engine is built to enable real-time processing of query results and generation of visual charts, which in turn generates structured reports, automating the entire "questioning - analysis - charting - reporting" process. S23. Innovative design of directional channel direct connection visualization module and "data sandbox" security architecture to ensure data security and analysis capabilities.

10. The solid-state hydrogen storage material design platform based on large language model and big data analysis according to claim 9, characterized in that: The implementation steps of the data sandbox in step S23 are: S231. The user question and database structure information are combined and used only to generate SQL commands, while the actual data does not pass through the external large model or is protected from unauthorized access; S232. All data analysis and visualization operations are performed locally to ensure that data is not leaked to the external environment; S233. Adopt security policies and access control mechanisms to ensure that only authorized users can access and process data in the data sandbox.

Images