Environmental data and quality information management system based on artificial intelligence

Abstract

The invention relates to an environmental data and quality information management system based on artificial intelligence, which integrates multi-source environmental data through a data collection engine, realizes standardized processing of unstructured data by using an artificial intelligence model, and constructs a total data table; a double-closed-loop data checking system is adopted to ensure the data accuracy; according to the method, comprehensive integration and standardized management of different data categories are achieved, the data tables can be used for subsequent data analysis and modeling, efficient support is provided for research on migration characteristics and hydrogeological conditions of pollutants, and a decision report can be automatically output in combination with a report generation module. According to the system, the environmental data management efficiency and the intelligent level can be remarkably improved, and accurate support is provided for pollution prevention and control.

Description

Technical Field

[0001] The present invention relates to the technical field of environmental monitoring and data management, and in particular to an environmental data and quality information management system based on artificial intelligence. Background Art

[0002] Taking soil and groundwater data as an example, the current survey plan is uploaded in PDF format, the coordinates of the points are manually entered, the laboratory test data is uploaded in PDF format, and the generated reports are also uploaded in PDF format. A large amount of data is stored in PDF files. The degree of digitization is low. A large number of photos on site and information data on drilling are pasted in documents as pictures and finally stored in PDF format. Therefore, it is particularly important to integrate, analyze and predict the data left over from multiple rounds of work guided by laws, regulations, ordinances and implementation plans on soil and groundwater, and to realize the systematic, complete and professional management and development and utilization of data in the direction of soil and groundwater.

[0003] It can be seen that there are the following challenges in the current ecological environment data management:

[0004] Complex data: There is a large amount of historical data sealed in various files, a wide variety of pollutants, inconsistent data formats and units, and scattered sources.

[0005] Low data processing efficiency: Manual cleaning and sorting of data is time-consuming and labor-intensive, with a high error rate.

[0006] Insufficient intelligence: Lack of in-depth analysis capabilities for environmental data makes it difficult to meet the needs of trend forecasting and intelligent decision-making. Summary of the invention

[0007] The purpose of the present invention is to propose an environmental data and quality information management system based on artificial intelligence in view of the prior art, which is used to solve the problems of scattered environmental monitoring data, inconsistent formats, low processing efficiency and insufficient intelligent analysis.

[0008] To achieve the above objectives, the present invention provides an artificial intelligence-based environmental data and quality information management system, comprising:

[0009] Data collection engine DGE, used to collect environmental monitoring data from multi-source heterogeneous data;

[0010] Real-time data management module, to obtain data from real-time data logger;

[0011] Sampling Plan Module SPM, used to design and manage sample collection plans and input planned sample information into the system;

[0012] The data processor and checker DP standardizes, integrates and verifies the data based on the artificial intelligence big model, and generates a total data table based on the set data form EQDD template; wherein, the data processor and checker DP recognizes the unstructured data in PDF, Excel, and pictures through the trained artificial intelligence big model, and converts it into standardized fields and organizes it into a unified format;

[0013] MongoDB database, which stores structured and unstructured environmental data, including the data quality management module DQM, which is maintained in a consistent format using data delivery DD, by performing standardized data input and processing, and also includes real-time monitoring of data in the real-time data management module;

[0014] Report generation module, which automatically outputs analysis reports based on the internalized artificial intelligence big model;

[0015] The double closed-loop data checking system, including the stand-alone version DP and the enterprise version DP, is used to verify the accuracy of data step by step, fill the required data into the data delivery DD, load the data of the data form into the data processing and checker DP, conduct a preliminary check of the format and content, check the data format, check, identify and record the data errors or inconsistencies in the data delivery DD through the data processing and checker, and then correct and adjust the data errors or inconsistencies found to verify the accuracy of the data. After the correction, the accuracy of the data is verified again through the data processing and checker DP, and the checked and corrected data form EQDD is loaded to complete the data loading process.

[0016] Preferably, the data collection engine DGE also includes integrating data from a laboratory information management system.

[0017] Preferably, the step of processing the data and unifying the format of the checker DP comprises:

[0018] By integrating the developed application programming interface API with Unstructured-IO's OCR, the content of various documents can be automatically identified and converted into structured data. Files in different formats can be converted into HTML or Markdown format, retaining the document structure. The text part is directly used for NLP preprocessing, and the image / table part is converted into structured data through OCR or table parsing tools.

[0019] Preferably, the data processor and checker DP verifies the data including: checking the standardization, missing, blank and / or corresponding errors of the table data and giving prompt feedback until the check is qualified.

[0020] Preferably, the double closed-loop data checking system comprises:

[0021] (1) Data provider submits to the stand-alone DP using the stand-alone DD form;

[0022] (2) Use the stand-alone DP system to check the correctness of the DD form. If the stand-alone DP system finds errors in the DD form, the stand-alone DP system will prompt and require the form to be revised and resubmitted until it meets the stand-alone DP system inspection requirements;

[0023] (3) The DD form checked by the stand-alone DP system is submitted to the enterprise DP;

[0024] (4) The enterprise version of the DP system further verifies the correctness of the DD form and conducts a more in-depth inspection. If the DD form is incorrect, the system will send a rejection notice and an error report to the data provider to request revision. The revised DD form is resubmitted for further inspection until it passes the monitoring.

[0025] (5) The DD form that has undergone double verification is submitted to the MongoDB database to complete the final storage of the data and send an acceptance notification.

[0026] Preferably, it also includes a three-dimensional visualization module DZV for generating pollution distribution maps and hydrogeological models. The three-dimensional visualization module DZV combines hydrogeological data with pollutant concentrations to generate a dynamic diffusion simulation map and mark restoration priority areas.

[0027] Preferably, the report generation module calls the artificial intelligence big model interface API to automatically generate risk assessment reports and decision support documents.

[0028] Preferably, the data delivery DD functionality includes access to a MongoDB database, an Excel spreadsheet, a tab-delimited text file, and a compressed file.

[0029] Preferably, the artificial intelligence big model includes the DeepSeek big model, the Tongyi Qianwen big model, the Grok xAI big model, and the ChatGPT big model.

[0030] Based on the above technical solution, the advantages of the present invention are:

[0031] The system of the present invention can construct ecological environment data tables in various fields, realize comprehensive integration and standardized management of different data categories, and these data tables can not only be used for subsequent data analysis and modeling, but also provide efficient support for studying the migration characteristics of pollutants and hydrogeological conditions. At the same time, the API based on the artificial intelligence big model is internalized into the AI-EQIS system, which can realize the generation of various complex reports at this level, and realize the integration, calling and sharing of various data and reports. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0033] Figure 1 : The overall system architecture diagram shows the interaction between each module;

[0034] Figure 2 : Stand-alone DP data processing flow;

[0035] Figure 3 : Enterprise Edition DP data processing flow;

[0036] Figure 4 : Double closed-loop data check flow chart;

[0037] Figure 5 : Data form EQDD master list example. DETAILED DESCRIPTION

[0038] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments.

[0039] The present invention provides an artificial intelligence-based environmental data and quality information management system, such as Figure 1~Figure 4 As shown, a preferred embodiment of the present invention is shown.

[0040] For the convenience of description, various English abbreviations involved in the present invention are as follows:

[0041] Data collection engine (DGE data Gather Engine), data delivery (DD data deliverable), data form (EQDD Environmental quality data deal) data processing and checker (DP Dataprocessor), sampling plan module (SPM sample planning Module), data quality management module (DQM dataquality module), three-dimensional data display module (DZV data Z-axis view), environmental quality information system information agent (EIA EQIS information agent), application programming interface (API ApplicationProgramming Interface), artificial intelligence-based environmental data and quality information management system (AI-EQIS).

[0042] like Figure 1 As shown, the AI-EQIS environmental data and quality information management system based on artificial intelligence includes:

[0043] The data collection engine DGE is used to collect environmental monitoring data from multi-source heterogeneous data. It is responsible for collecting data from different sources and providing basic information for subsequent processing. DGE is only used in a single-machine workflow process. In addition to collecting data from different sources, it can also integrate data from laboratory information management systems and real-time data loggers to ensure the comprehensiveness of the data.

[0044] Real-time data management module, obtains data from real-time data logger.

[0045] The sampling plan module SPM is used to design and manage sample collection plans, ensure the representativeness and effectiveness of samples, and input the planned sample information into the system to prepare for data collection and processing.

[0046] The data processor and checker DP standardizes, integrates and verifies data based on the artificial intelligence big model, and generates a total data table based on the set data form EQDD template. It is responsible for checking and verifying the collected data, and for ensuring that all EQDD forms input into the system at different stages meet the system's data requirements. The main means of implementation is to set the form template and use the artificial intelligence big model to recognize (it can be a professional model trained on the basis of the Deep seek big model) to quickly identify the form data and perform checks and verifications, verify the accuracy and completeness of the data, and ensure that the data quality meets the requirements.

[0047] MongoDB database, which stores structured and unstructured environment data, includes the data quality management module DQM, which uses data delivery DD to maintain it in a consistent format, performs standardized data input and processing, and also includes real-time monitoring of data in the real-time data management module.

[0048] The report generation module automatically outputs analysis reports based on the internalized artificial intelligence big model.

[0049] The double closed-loop data checking system, including the stand-alone version DP and the enterprise version DP, is used to verify the accuracy of data step by step, fill the required data into the data delivery DD, load the data of the data form into the data processing and checker DP, conduct a preliminary check of the format and content, check the data format, check, identify and record the data errors or inconsistencies in the data delivery DD through the data processing and checker, and then correct and adjust the data errors or inconsistencies found to verify the accuracy of the data. After the correction, the accuracy of the data is verified again through the data processing and checker DP, and the checked and corrected data form EQDD is loaded to complete the data loading process.

[0050] The artificial intelligence-based environmental data and quality information management system AI-EQIS of the present invention mainly includes three levels of workflow. The first process is data collection and processing, which mainly includes artificial intelligence-assisted data recognition and organization, data collection and data processing engine, laboratory data and management, sampling plan module, real-time data management (mainly including the collection and organization of online monitoring data) and data processing and inspection.

[0051] When the AI-EQIS system standardizes and integrates data based on the artificial intelligence big model, the artificial intelligence big model has collected industry laws, standards, specifications, guidelines, instructions, letters and other documents of the ecological environment system through artificial intelligence technology as support, and has trained one or more of the selected big models (DeepSeek big model, Tongyi Qianwen big model, Grok xAI big model, ChatGPT big model) with environmental impact assessment texts, investigation reports, risk assessment reports, technical plans, supervision reports, effect evaluation reports, acceptance reports, process design plans, feasibility study reports, three-dimensional simulation methods for soil and groundwater pollution, atmospheric pollution simulation methods, watershed water pollution and water environment simulation methods, etc., and has been tested and used to meet the needs.

[0052] Since soil and groundwater pollution data are more complex and involve more data types and formats, in order to further illustrate the data collection, inspection, and upload mode of the system of the present invention, the present invention takes soil and groundwater pollution as an example. The data summary tables generated at this time include soil pollution summary data table, groundwater pollution summary data table, and hydrogeological summary data table, as follows:

[0053] The present invention uses the published artificial intelligence big model as a tool to identify and organize previously stored data of various categories (which can be various image formats, PDF formats, Excel formats, etc.) into a unified format (which can be Excel format or other formats). The above-mentioned identified stock data is mainly organized into two categories of data: soil and groundwater pollution detection data (pollutant data) and hydrogeological data;

[0054] Specifically, the pollutant data integration method is: the identified soil pollutant categories and groundwater pollutant categories are integrated into two parallel databases, namely the "soil pollution total data table" and the "groundwater pollution total data table". The core fields of the total data table include: pollutant category, pollutant concentration unit, sampling time, and classification information based on the nature of pollutants, such as heavy metal pollutants, volatile organic compounds (VOCs), semi-volatile organic compounds (SVOCs), new pollutants and persistent organic pollutants (POPs). In the process of building the total data table, the pollutant identification model formed through training will be further used for relevant information in the target area to automatically extract information such as pollutant category, concentration, time, etc., and integrate it into the corresponding total data table.

[0055] Furthermore, the hydrogeological data integration method is similar to the pollutant data integration method to establish an independent "hydrogeological total data table" for storing and managing hydrogeological related information. The main fields of the hydrogeological total data table include: sampling point coordinates, ground elevation, groundwater depth, lithology information, geotechnical test data, pumping test data, and well completion report summary. In the process of building the total data table, the key information is automatically extracted from the target data (such as well completion reports or sampling records in PDF format) through the training and optimization of the hydrogeological data recognition model, and standardized integration is achieved.

[0056] Therefore, through the above methods, soil pollution data tables, groundwater pollution data tables and hydrogeological data tables are constructed respectively to achieve comprehensive integration and standardized management of different data categories. These data tables can not only be used for subsequent data analysis and modeling, but also provide efficient support for studying the migration characteristics of pollutants and hydrogeological conditions.

[0057] It should be noted that the above-mentioned soil and groundwater pollution data collection and analysis methods can also be used in the fields of atmosphere, solid waste, sewage treatment plants, mining environment, water ecology and water environment, ecological survey and biodiversity. The EQDD form template is mainly designed according to the different subdivisions of the ecological environment system, and the data is processed through the above-mentioned process of processing soil and groundwater pollution data through data collection, inspection and upload.

[0058] As for the stock data, the present invention uses a large artificial intelligence model to assist in identification and integration. For the data that is currently occurring, an interface can be directly set up to organize and store it according to a standard data table template.

[0059] The data processor and checker (DP) is mainly responsible for checking the data before various links enter AI-EQIS. It mainly checks the standardization, missing, blank, correspondence and other errors of the table data and gives prompt feedback. Only after the inspection is passed can it be linked to AI-EQIS.

[0060] The second part of the AI-EQIS workflow of the present invention is the work interface, which mainly includes a professional version interface system and an enterprise version interface system. The professional version interface system is mainly used by scientific and technological professionals and AI-EQIS system staff to check the quality of input data and import high-quality data into the system. The enterprise version interface system mainly serves enterprises and institutions using the AI-EQIS system to import and export data, and the enterprise version interface system is also linked to various GIS databases, "CAS substance library" (CIRS Group database) and other databases, and supports connections with GIS systems, third-party tools such as MATLAB and "CAS substance library" (CIRS Group database).

[0061] The third part of the AI-EQIS workflow is the in-depth analysis and secondary development of the database, which mainly includes:

[0062] (1) Based on the professional model of artificial intelligence big model, data can be analyzed, mainly including the reliability and change trend analysis of multiple rounds of pollution data and hydrological data of the same enterprise, park, and plot. It can be expected to give an analysis of whether the data is reliable, the degree of reliability, and the degree of abnormality. Different abnormalities give different early warning prompts. And based on the systematic analysis of the above long-term and multiple rounds of data, a prediction of pollution spread is given. The above data analysis is mainly displayed in the format of icons. The results of the above data analysis are used to support the supervision of key areas and key parks (Class I chemical parks).

[0063] (2) Based on the collected and organized soil and groundwater pollution data (including complete hydrogeological data), combined with certain algorithms and software interfaces, three-dimensional hydrogeological maps and three-dimensional pollution distribution maps can be directly produced. Based on the generated intuitive and systematic three-dimensional pollution distribution maps and three-dimensional hydrogeological maps, technical solutions for the remediation and control of contaminated soil and groundwater can be accurately developed.

[0064] (3) Based on systematic and complete data analysis, scientific pollution source analysis can be provided, providing a basis for systematic and accurate risk assessment (such as human health risk assessment and environmental risk assessment). In the past, risk assessments were based on weak data, mainly based on one-time data and small plots of land. However, the present invention can calculate the pollution carrying capacity of soil and groundwater based on systematic and complete data, and provide policy and decision-making basis for other demand parties. At the same time, the API based on the Deep Seek large model can be developed and internalized into the AI-EQIS system, which can make the generation of various complex reports a reality at this level, and can realize the integration, calling and sharing of various data and reports.

[0065] like Figure 2 , Figure 3 As shown, the data processing flow of the stand-alone DP and the enterprise DP of the present invention is as follows:

[0066] Fill the required data into the data delivery DD to ensure the completeness and accuracy of the data. Load the data form EQDD data into the data processing and checker DP, conduct a preliminary check of the format and content, check the data format, and identify and record data errors or inconsistencies in the data delivery DD through the data processing and checker DP. Receive the original EQDD data file from the data provider and prepare for subsequent processing. Load the EQDD data into the data processing and checker DP, conduct a preliminary check of the format and content, identify data errors, and identify and record data errors or inconsistencies in the data delivery DD through the data processing and checker DP. Correct data errors. For data errors found in the data processing and checker DP check, make necessary corrections and adjustments to verify data accuracy. After correction, verify the accuracy of the data again through the data processing and checker DP to ensure that there are no errors. Submit the final data and load the checked and corrected data form EQDD into AI-EQIS to complete the data loading process.

[0067] Further, if Figure 4 As shown, the present invention adopts a double closed-loop data inspection system to form a local inspection closed loop to ensure that the data is strictly inspected before submission: The specific path is as follows:

[0068] (1) Data provider submits to the stand-alone DP using the stand-alone DD form;

[0069] (2) Use the stand-alone DP system to check the correctness of the DD form. If the stand-alone DP system finds errors in the DD form, the stand-alone DP system will prompt and require the form to be revised and resubmitted until it meets the stand-alone DP system inspection requirements;

[0070] (3) The DD form checked by the stand-alone DP system is submitted to the enterprise DP;

[0071] (4) The enterprise version of the DP system further verifies the correctness of the DD form and conducts a more in-depth inspection. If the DD form is incorrect, the system will send a rejection notice and an error report to the data provider to request revision. The revised DD form is resubmitted for further inspection until it passes the monitoring.

[0072] (5) The double-verified DD form is submitted to the MongoDB database to complete the final storage of the data. The system will send an acceptance notification indicating that the data has passed all checks.

[0073] The above-mentioned double closed-loop data checking system can ensure that the data also undergoes a closed-loop check during the submission process, further improving the accuracy and reliability of the data.

[0074] like Figure 5 As shown, an example of a data form EQDD master form is given. The DD form can be a variety of forms (such as point basic information form, drilling engineering quantity form, monitoring well information form, well construction data form, soil layer information form, water level information form, water quality quick screening form, geotechnical experiment form, field sampling situation form, soil sample parameter form, etc.), each form displays different content, and the total form displays different tabs. Preferably, colors can be used as a distinction in the form, such as: red + horizontal bar, it is required to fill in the content; red is required to fill in the content; blue is very important content; yellow highlights are DP inspection content parts.

[0075] Data Delivery DD can access databases for data retrieval and management, providing a structured approach to data processing. Data Delivery DD is compatible with Excel spreadsheets, allowing easy entry and manipulation of data in a format familiar to users. Data Delivery DD is also compatible with tab-delimited text files, Data Delivery DD can utilize tab-delimited text files, providing a simple and effective way to organize and store data.

[0076] In the AI-EQIS database, the data can be further managed and analyzed. The data delivery DD can enter a database, an Excel spreadsheet, a tab-delimited text file or a compressed file. Further features of the data delivery DD of the present invention are as follows:

[0077] Data Compression, Data Delivery DD can handle compressed files, helping to reduce file size and improve data transfer efficiency.

[0078] Data integrity, using compressed files ensures data integrity during compression and decompression, maintaining accuracy.

[0079] Cross-platform compatibility: compressed files are universally compatible across different platforms, ensuring seamless data sharing.

[0080] Since a large amount of soil and groundwater data is currently submitted in the form of PDF text, the data involved in the PDF text needs to be identified, extracted and classified through artificial intelligence methods. Since a large number of pictures, GIS vector graphics and other complex types of data need to be identified in the PDF text at the same time, it is difficult to fully identify all the data.

[0081] The specific implementation method of the present invention based on the artificial intelligence large model to assist in the recognition and collection of historical data is as follows:

[0082] Unified multi-format document processing, through the integration of developed API and Unstructured-IO's OCR, automatically identify various documents (Word, Excel, PPT, PDF, pictures, etc.) and convert them into structured data for subsequent data calls and report generation calls; format unification, convert different format files (Word / PPT / PDF, etc.) into HTML or Markdown format, retaining the document structure (chapter, paragraph, style, etc.). In the content separation and processing process, the text part is directly used for NLP preprocessing, and the picture / table part is converted into structured data (such as text or JSON) through OCR or table parsing tools.

[0083] Structured output, which retains the document's hierarchical structure, text, charts and other information, and stores them in a unified format (such as JSON) to ensure data relevance and scalability;

[0084] Automatic classification, automatic identification of file type and content type (text / table / image), and calling of corresponding processing modules.

[0085] Special processing uses the best method in conventional ways for different content types (such as OCR for images and table parsing tools for Excel). For complex formats (such as PPT), tool chains are used to solve compatibility issues.

[0086] The data integration process integrates the processed content into a unified JSON format, including metadata and related information.

[0087] The AI ​​large model training data set generation process requires data segmentation to achieve data set training. The training set, validation set, and test set are divided in proportion (80 / 10 / 10 by default), token restrictions are set, and the text block length is controlled (100-500 tokens) to adapt to the model input requirements.

[0088] By using an AI-assisted large-scale model, the data structure can be retained, and the original hierarchy and style of the document can be maintained through HTML / Markdown conversion. Data can be processed with flexibility and precision, and text, images, and tables can be processed by type to improve the accuracy of subsequent processing. The entire process does not require human intervention, achieving automation and efficiency, and can quickly process historical data piled up in PDF documents in large quantities.

[0089] During the development process, it is necessary to integrate a variety of parsing tools and scripts (such as doc to HTML, PDF to Markdown). The database identified and constructed by the above method can directly retrieve data and automatically identify it to the parent order ( Figure 5The EQDD form in the example above can automatically generate a time series form after training for subsequent call analysis. The automatically generated EQDD form is submitted to the DP module, which automatically analyzes the comprehensiveness and completeness of the data. After passing the inspection, it enters the AI-EQIS system to complete the identification and collection of historical data.

[0090] In order to uniformly store and process and analyze data from files in different formats (such as Word, Excel, PPT, JPG, PDF, TIF), the present invention uses MongoDB as a non-relational database because it supports flexible data structures.

[0091] Design a document-based data model, convert each file type into a corresponding MongoDB document, and common formats of data models include the following:

[0092] Word / PPT: text object or main document contains embedded objects (table / chart);

[0093] Excel: Converted to JSON format, including rows, columns and metadata information;

[0094] JPG / TIF: Analyzes the image’s pixel data, color information, etc., and stores them as image documents.

[0095] Data extraction and conversion: Use Python toolkits such as Pandas to process Excel data and convert the content into DataFrame or CSV files. Extract text and embedded objects (tables / charts) from Word / PPT. Image files are parsed using the Pillow library and image data is stored in MongoDB.

[0096] Data integration includes: using ETL tools to convert multi-source data into a unified format, such as JSON or CSV, and then inserting it into MongoDB. Or directly mapping the data structure of each file type to a MongoDB document to ensure data integrity and consistency.

[0097] Data processing and analysis include: using aggregate queries in MongoDB databases for complex data manipulation and analysis. Using Python to write scripts for data cleaning, transformation, and statistical analysis, combined with visualization tools such as Matplotlib or Seaborn.

[0098] Performance optimization and expansion include: Consider creating indexes to improve query performance as needed. Compress or degrade image data to reduce storage space and transmission time.

[0099] Middleware and Automation: Use Elasticsearch or other search engines to quickly retrieve data from MongoDB databases. Write automated scripts (such as Shell or Python) to handle file conversion and data extraction processes to ensure efficient operation.

[0100] Through the above steps, unified storage, processing and analysis of data in different types of files can be achieved, meeting different needs while improving efficiency.

[0101] The present invention performs data quality management through the data quality management module DQM to ensure the reliability of data. The data organization of the data quality management module DQM is guaranteed by the following aspects:

[0102] Data delivery DD helps in organizing data in a structured manner for easy analysis and reporting. Using data delivery DD, data is maintained in a consistent format which is essential for accurate data interpretation.

[0103] Data delivery (DD) reduces errors in data management by enforcing standardized data entry and processing methods.

[0104] The data can be reviewed through the professional version interface system and combined with the internalization function of the artificial intelligence model to ensure data quality.

[0105] Since most other fields such as atmosphere, sewage, watershed water environment, and solid waste have achieved online monitoring and networking management, the online data and stored data can be obtained relatively directly, and the interface can be directly developed to connect to the AI-EQIS system to integrate the data for digital use. Therefore, in addition to the data collected by the AI-EQIS system and uniformly classified into the database, the real-time online data accessed is also within the professional data management series, using the AI-EQUIS real-time data management module for real-time data monitoring to promptly discover and solve problems.

[0106] Furthermore, based on the system of the present invention, data analysis, report generation and data sharing collaboration can be realized, and more functional expansion can be achieved.

[0107] Leveraging data to deliver DD can enhance data insights, and users can gain deeper insights into data by leveraging advanced analytical tools and techniques.

[0108] Data Delivery (DD) enables efficient processing of large data sets, improving the speed and accuracy of data analysis.

[0109] Data Delivery (DD) supports the decision-making process by providing reliable and timely data to enable informed choices.

[0110] Use the AI-EQIS Enterprise Edition interface to conduct in-depth data analysis to support decision making. The Professional Edition AI-EQIS Portal can also perform comprehensive data analysis and generate various charts, texts, 3D graphs and reports.

[0111] The system includes a report generation module, which automatically outputs analysis reports based on the internalized artificial intelligence large model, and also includes a three-dimensional visualization module DZV, which is used to generate pollution distribution maps and hydrogeological models. The three-dimensional visualization module DZV combines hydrogeological data with pollutant concentrations to generate dynamic diffusion simulation maps and mark remediation priority areas.

[0112] Specifically, both the professional version interface and the enterprise version interface have the opportunity to fully use the database data when entering the system, and the enterprise version interface system can connect to the three-party interface, connect various geographic information data, data from other monitoring systems, etc. At the same time, the enterprise version interface system can also develop various plug-ins, dashboards, and visualization tools, use EIAs, DZV, EIA and APIs of regulatory departments at all levels for data visualization, and intuitively display analysis results.

[0113] The enterprise version system integrates with other systems through the AI-EQIS third-party interface to expand data content, capacity and processing capabilities. It can widely integrate ecological environmental data with data from other departments to meet various complex needs.

[0114] Use dashboards and widgets to display key data indicators for easy monitoring and decision-making.

[0115] When the enterprise version system is connected with third-party interfaces, AI-EQIS allows spatial data analysis and visualization, and the data is delivered to DD-compatible statistical software, supporting comprehensive data analysis and modeling.

[0116] The AI-EQIS system of the present invention can promote data sharing, and data delivery DD promotes data sharing among different stakeholders, enhancing collaboration and transparency. By using the AI-EQIS system, teams can collaborate more effectively on data projects, thereby achieving better results. The AI-EQIS system supports remote work, and data delivery DD supports remote work, allowing data to be accessed from anywhere, promoting flexibility and productivity, and being able to support cloud computing. And because the functions of the Deep Seek big model and other models (such as the Tongyi Qianwen big model, the Grok xAI big model, and the ChatGPT big model) are internalized, it can be more efficient and accurate when generating texts such as report generation, industry risk assessment, and environmental assessment reports, and can achieve accurate and rapid generation of various charts.

[0117] Furthermore, through the system of the present invention, after integrating the data collection of the entire environmental field into the AI-EQIS system, the data can be efficiently called, and comprehensive pollution data, water resource data, etc. can be called according to the coordinate range, and industrial system data (such as enterprise coordinates, production raw materials and auxiliary materials, past environmental impact assessment data, sewage discharge data, etc.), natural resource system data (such as three-dimensional GIS maps, woodlands, mines, water sources, various land boundaries, etc.), and agricultural system data (non-point source pollution, rural sewage discharge and treatment, etc.) can be accessed. In this way, the pollution situation can be analyzed comprehensively, quickly and efficiently, and the pollution trend can be predicted.

[0118] The system of the present invention can construct the total data table of each ecological environment segment respectively, realize the comprehensive integration and standardized management of different data categories, and these total data tables can not only be used for subsequent data analysis and modeling, but also provide efficient support for studying the migration characteristics of pollutants and hydrogeological conditions. At the same time, the API based on the artificial intelligence big model is internalized into the AI-EQIS system, which can realize the generation of various complex reports at this level, and realize the integration, calling and sharing of various data and reports.

[0119] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or some technical features can be replaced by equivalents without departing from the spirit of the technical solution of the present invention, which should be included in the scope of the technical solution for protection of the present invention.

Claims

1. An artificial intelligence-based environmental data and quality information management system, characterized in that: include: Data collection engine DGE, used to collect environmental monitoring data from multi-source heterogeneous data; Real-time data management module, to obtain data from real-time data logger; Sampling Plan Module SPM, used to design and manage sample collection plans and input planned sample information into the system; The data processor and checker DP standardizes, integrates and verifies the data based on the artificial intelligence big model, and generates a total data table based on the set data form EQDD template; wherein, the data processor and checker DP recognizes the unstructured data in PDF, Excel, and pictures through the trained artificial intelligence big model, and converts it into standardized fields and organizes it into a unified format; MongoDB database, which stores structured and unstructured environmental data, including the data quality management module DQM, which is maintained in a consistent format using data delivery DD, by performing standardized data input and processing, and also includes real-time monitoring of data in the real-time data management module; Report generation module, which automatically outputs analysis reports based on the internalized artificial intelligence big model; The double closed-loop data checking system, including the stand-alone version DP and the enterprise version DP, is used to verify the accuracy of data step by step, fill the required data into the data delivery DD, load the data of the data form into the data processing and checker DP, conduct a preliminary check of the format and content, check the data format, check, identify and record the data errors or inconsistencies in the data delivery DD through the data processing and checker, and then correct and adjust the data errors or inconsistencies found to verify the accuracy of the data. After the correction, the accuracy of the data is verified again through the data processing and checker DP, and the checked and corrected data form EQDD is loaded to complete the data loading process.

2. The system according to claim 1, characterized in that: The data collection engine DGE also includes integrating data from laboratory information management systems.

3. The system according to claim 1, characterized in that: The steps of processing the data and unifying the format of the inspector DP include: By integrating the developed application programming interface API with Unstructured-IO's OCR, the content of various documents can be automatically identified and converted into structured data. Files in different formats can be converted into HTML or Markdown format while retaining the document structure. The text part can be directly used for NLP preprocessing, and the image / table part can be converted into structured data through OCR or table parsing tools.

4. The system according to claim 1, characterized in that: The data processor and checker DP verifies the data by checking the standardization, missing, blank and / or corresponding errors of the table data and giving prompt feedback until the check is qualified.

5. The system according to claim 1, characterized in that: The data checking steps of the double closed-loop data checking system include: (1) Data provider submits to the stand-alone DP using the stand-alone DD form; (2) Use the stand-alone DP system to check the correctness of the DD form. If the stand-alone DP system finds errors in the DD form, the stand-alone DP system will prompt and require the form to be revised and resubmitted until it meets the stand-alone DP system inspection requirements; (3) The DD form checked by the stand-alone DP system is submitted to the enterprise DP; (4) The enterprise version of the DP system further verifies the correctness of the DD form and conducts a more in-depth inspection. If the DD form is incorrect, the system will send a rejection notice and an error report to the data provider to request revision. The revised DD form is resubmitted for further inspection until it passes the monitoring. (5) The DD form that has undergone double verification is submitted to the MongoDB database to complete the final storage of the data and send an acceptance notification.

6. The system according to claim 1, characterized in that: It also includes a three-dimensional visualization module DZV for generating pollution distribution maps and hydrogeological models. The three-dimensional visualization module DZV combines hydrogeological data with pollutant concentrations to generate dynamic diffusion simulation maps and mark restoration priority areas.

7. The system according to claim 1, characterized in that: The report generation module calls the artificial intelligence big model interface API to automatically generate risk assessment reports and decision support documents.

8. The system according to claim 1, characterized in that: The data delivery (DD) functionality includes access to MongoDB databases, Excel spreadsheets, tab-delimited text files, and compressed files.

9. The system according to claim 1, characterized in that: The data summary tables include a soil pollution summary data table, a groundwater pollution summary data table, and a hydrogeological summary data table.

Images