Production safety inspection system and method in secret-related offline environment

Abstract

The invention discloses a production safety inspection system and method in a secret-related off-line environment, and relates to the technical field of safety production, and the system comprises a basic information module which stores basic information, and the basic information comprises place information and personnel information; the form management module is used for segmenting old form data into specific inspection contents and classifying the inspection contents into inspection items to obtain a self-defined form; the association module is used for establishing and storing an association relationship between the basic information and a custom form; the form filling module is used for calling a corresponding self-defined form according to the basic information and the association relationship and carrying out checking, information correction and information filling to generate result data; and the result transmission module transmits the result data offline. According to the method, existing forms are scattered, recombined according to rules and associated with basic information such as personnel, places and the like, so that self-definition of the security check forms is realized, check and data collection are quickly completed, and the check efficiency and the information accuracy are improved.

Description

Technical Field

[0001] This invention relates to the field of safety production technology, specifically, to a production safety inspection system and method for a classified offline environment. Background Technology

[0002] Currently, with the development of technology and the increasing demands for production safety, the level of informatization in production safety inspections is constantly improving. In most cases, safety inspections using networks and terminals are already widespread. However, security inspection sites in military or classified units are often in physically isolated environments, where wireless communication functions of equipment are not permitted. Traditional methods using paper-based security inspection records suffer from inefficiency, susceptibility to errors and omissions, and low levels of structure. Furthermore, due to the wide variety and highly specialized nature of security inspection forms, paper-based security inspection records cannot meet the needs of military or classified units for efficient inspections and accurate information collection in offline environments. Summary of the Invention

[0003] The purpose of this invention is to provide a production safety inspection system and method for a classified offline environment, which solves the problems of low efficiency, low structuring degree, and inability to meet the requirements of accurate information collection in the prior art using paper-based safety inspection record forms.

[0004] The present invention solves the above problems through the following technical solution:

[0005] A production safety inspection system for classified offline environments includes:

[0006] The basic information module is configured to store basic information, including location information and personnel information.

[0007] The form management module is configured to split old form data into specific inspection contents and categorize the inspection contents into inspection items to obtain custom forms;

[0008] The association module is configured to establish and store the association between the basic information and the custom form;

[0009] The form filling module is configured to retrieve the corresponding custom form based on the basic information and relationships, and perform checks, information corrections and information filling to generate result data.

[0010] The result transmission module is configured to transmit the result data offline.

[0011] Furthermore, the form management module includes:

[0012] The data model design unit is configured to define the data structures of the old form data table old_form_data, the check item table check_item, and the check content table check_content.

[0013] The old data parsing unit is configured to retrieve the original form content and form type of the target form by querying the old form data table; and to select the corresponding parsing strategy according to the form type to perform structured parsing to obtain key fields or key-value pairs parsedData.

[0014] The content segmentation unit is configured to split the original content into the smallest granularity of inspection content based on the parsed key-value pairs parsedData, or to map it to preset inspection content in combination with business rules.

[0015] The classification and matching unit is configured to establish a matching rule base. The matching rule base has a preset mapping relationship between the key name of the check content and the check item. It matches the key-value pair parsedData with the corresponding check item ID item_id. If the match is successful, a CheckContent object is generated. The CheckContent object is associated with the unique form ID form_id, the check item ID item_id, and the specific value content_value. Otherwise, manual intervention is triggered. The unit also provides feedback on the processing results and automatically updates the check content table check_content after manual correction, and synchronizes the new mapping relationship to the matching rule base.

[0016] The front-end interface is configured to receive and display the processing results, allowing users to manually select corresponding inspection items, correct unclassified content, and add inspection items and content, and then feed back to the classification and matching unit.

[0017] Furthermore, the old form data table old_form_data is used to store the original form data, and its fields include the unique form identifier form_id, the original form content raw_content, and the form type form_type.

[0018] Furthermore, the check_item table is used to store standardized check items, and its fields include check item ID (item_id), check content (content), and check method (method).

[0019] Furthermore, the check_content table is used to store the specific content extracted from the original form data. The fields include content ID content_id, associated form unique identifier form_id, associated check item ID item_id, specific value content_value, and additional information ext_info.

[0020] Furthermore, the processing results include successfully categorized inspection content, the number of uncategorized content, and manually corrected uncategorized content.

[0021] Furthermore, the result data is a structured detection result or a QR code of a single result. The result data is transmitted offline via USB or by scanning a QR code to a standalone or isolated environment.

[0022] Furthermore, it also includes a QR code recognition module and a pattern lock module. The QR code recognition module is configured to establish a unique association between the location and the QR code by encoding the location information into the QR code, and store it in the basic information module. The pattern lock module is configured to establish a unique association between the pattern lock and personnel information by storing the information drawn by the pattern lock in a string, and store it in the basic information module.

[0023] A method for production safety inspection in a classified offline environment includes:

[0024] Store basic information, including location information and personnel information;

[0025] The old form data is divided into specific inspection items, and the inspection items are categorized into inspection items to obtain a custom form;

[0026] Establish and store the association between the basic information and the custom form;

[0027] Based on the basic information and relationships, retrieve the corresponding custom form and perform checks, information corrections, and information filling to generate result data;

[0028] The resulting data will be transmitted offline.

[0029] Furthermore, the process of segmenting the old form data into specific inspection contents and categorizing these contents into inspection items to obtain a custom form specifically includes:

[0030] Define the data structures for the old form data table `old_form_data`, the check item table `check_item`, and the check content table `check_content`. The old form data table `old_form_data` stores the original form data, with fields including the unique form identifier `form_id`, the original form content `raw_content`, and the form type `form_type`. The check item table `check_item` stores standardized check items, with fields including the check item ID `item_id`, the check content `content`, and the check method `method`. The check content table `check_content` stores the specific content extracted from the original form data, with fields including the content ID `content_id`, the associated unique form identifier `form_id`, the associated check item ID `item_id`, the specific value `content_value`, and additional information `ext_info`.

[0031] By querying the old form data table, the original form content and form type of the target form can be obtained; and the corresponding parsing strategy can be selected according to the form type to perform structured parsing to obtain key fields or key-value pairs parsedData;

[0032] Based on the parsed key-value pairs, the original content is split into the smallest granularity of inspection content, or mapped to preset inspection content in combination with business rules;

[0033] Based on the pre-defined mapping relationship between check content keys and check items in the matching rule base, the key-value pair `parsedData` is matched against the corresponding check item ID `item_id`. If the match is successful, a `CheckContent` object is generated, which is associated with the unique form ID `form_id`, the check item ID `item_id`, and the specific value `content_value`. Otherwise, manual intervention is triggered, and the processing results are fed back, including the successfully categorized check content, the number of uncategorized content, and the manual correction of uncategorized content.

[0034] The system receives and displays the processing results, allowing users to manually select corresponding check items, correct unclassified content, and add new check items and content. After manual correction, the system automatically updates the check content table (check_content) and synchronizes the new mapping relationship to the matching rule base.

[0035] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0036] (1) This invention breaks down existing forms and reassembles them according to rules, and associates them with basic information such as personnel and locations, so as to realize the customization of security inspection forms and quickly complete inspections and data collection, thereby improving inspection efficiency and information accuracy.

[0037] (2) The present invention ultimately imports the form information into the isolated network via USB or QR code scanning without using wireless communication, thereby realizing safe production inspection in a classified offline environment and providing efficient and accurate basic data support for production safety management.

[0038] (3) This invention utilizes technologies such as form construction, QR code, and image lock to realize custom forms for security checks, rapid information filling, and rapid collection. Attached Figure Description

[0039] Figure 1 This is a system principle block diagram according to an embodiment of the present invention;

[0040] Figure 2 This is a flowchart of an embodiment of the present invention. Detailed Implementation

[0041] The present invention will be further described in detail below with reference to embodiments, but the implementation of the present invention is not limited thereto.

[0042] Example 1:

[0043] Combined with appendix Figure 1 As shown, a production safety inspection system for a classified offline environment includes:

[0044] The basic information module is configured to store basic information, including venue information and personnel information. Specifically, the venue information can be asymmetrically encrypted (RSA) and encoded into a QR code using a QR code recognition module to establish a unique association between the venue information and the QR code, and then stored in the basic information module. Alternatively, the graphic lock module can store the information drawn by the graphic lock in a string to establish a unique association between the graphic lock and the personnel information, and then store it in the basic information module.

[0045] The form management module is configured to split old form data into specific inspection contents and categorize the inspection contents into inspection items to obtain custom forms;

[0046] The association module is configured to establish and store the association between the basic information and the custom form;

[0047] The form filling module is configured to retrieve the corresponding custom form based on the basic information and association, and perform checks, information corrections and information filling to generate result data; wherein, the result data is a structured detection result or a QR code of a single result;

[0048] The result transmission module is configured to transmit the result data offline. It includes a USB transmission module and a QR code transmission module, enabling offline transmission of the result data via USB to a standalone or isolated environment, or by scanning a QR code to transmit the result data to a standalone receiving module or an isolated environment.

[0049] Furthermore, the form management module includes:

[0050] The data model design unit is configured to define the data structures of the old form data table `old_form_data`, the check item table `check_item`, and the check content table `check_content`. The old form data table `old_form_data` stores the original form data, with fields including the form unique identifier `form_id`, the original form content `raw_content`, and the form type `form_type`. The check item table `check_item` stores standardized check items, with fields including the check item ID `item_id`, the check content `content`, and the check method `method`. The check content table `check_content` stores the specific content extracted from the original form data, with fields including the content ID `content_id`, the associated form unique identifier `form_id`, the associated check item ID `item_id`, the specific value `content_value`, and additional information `ext_info`.

[0051] The old data parsing unit is configured to retrieve the original form content and form type of the target form by querying the old form data table; and to select the corresponding parsing strategy according to the form type to perform structured parsing to obtain key fields or key-value pairs parsedData.

[0052] The content segmentation unit is configured to split the original content into the smallest granularity of inspection content based on the parsed key-value pairs parsedData, or to map it to preset inspection content in combination with business rules.

[0053] The classification and matching unit is configured to establish a matching rule base. The matching rule base has a preset mapping relationship between the key name of the check content and the check item. It matches the key-value pair parsedData with the corresponding check item ID item_id. If the match is successful, a CheckContent object is generated. The CheckContent object is associated with the unique form ID form_id, the check item ID item_id, and the specific value content_value. Otherwise, manual intervention is triggered. The unit also provides feedback on the processing results (including successfully classified check content, the number of unclassified content, and manual correction of unclassified content, etc.), and automatically updates the check content table check_content after manual correction and synchronizes the new mapping relationship to the matching rule base.

[0054] The front-end interface is configured to receive and display the processing results, allowing users to manually select corresponding inspection items, correct unclassified content, and add inspection items and content, and then feed back to the classification and matching unit.

[0055] The form management module is responsible for breaking down and reorganizing forms to achieve form customization. Form customization adopts a template engine + dynamic parsing approach, providing basic form templates and allowing the extension of field logic and styles through code (such as HTML / JS). The backend dynamically generates forms by parsing template syntax (such as Vue / React components, Thymeleaf), balancing flexibility and development efficiency.

[0056] The underlying design of the custom form template is as follows:

[0057] ① Design the basic form template structure

[0058] Define the core elements of the template:

[0059] The template must include the basic framework of the form (such as layout containers, field placeholders, and interaction triggers) and an expandable area (for users to customize fields, styles, and logic).

[0060] Define template syntax rules:

[0061] Use specific identifiers (such as {{}}, {@}) to distinguish between "fixed content" and "dynamically expanded content," supporting logic such as field definition, condition judgment, and loops. For example: {{field:fieldType}}: indicates inserting a field of a specified type (such as {{field:select}} representing a dropdown list); {@if condition}...{@ / if}: indicates conditional rendering (such as displaying different fields based on user roles).

[0062] ② Provide user-defined interfaces:

[0063] Allow template-based expansion

[0064] Users can add custom fields, styles, or interactive logic to the expandable areas of the template through code (such as HTML, JS) or a visual editor. For example: Adding a new field: {{field:text, name:"safe", label:"Safety",required:true}} Custom style: <style>.custom-field { color: red;}< / style> Interactive logic: {@on change="handleChange()"} (binding the field change event).

[0065] Verify the legality of custom content:

[0066] The system verifies whether the custom content entered by the user conforms to the template rules (such as whether the field type is supported and whether the event name exists) through a syntax parser, to avoid rendering errors caused by illegal configurations.

[0067] ③ Back-end template parsing and data processing:

[0068] Parse the template and generate an Abstract Syntax Tree (AST):

[0069] After receiving the custom template submitted by the user, the back-end parses the syntax identifiers through a template engine (such as vue-loader for Vue or Thymeleaf for Java), and converts the template into an AST (Abstract Syntax Tree), clarifying the structure of fixed content, dynamic fields, and logical branches.

[0070] For example, when parsing {{field:select,options:["safe","dangerous"]}}, the AST will record that "the field type is a dropdown box, and the options are safe and dangerous".

[0071] Fill dynamic content in combination with business data:

[0072] If the template contains dynamic data (such as dropdown box options obtained from a database), the back-end reads the data from the business system (such as a database, API), replaces the placeholders in the AST, and generates the final executable form structure data (usually JSON).

[0073] ④ Front-end dynamic form rendering:

[0074] Receive the back-end parsing result:

[0075] The front-end obtains the form structure data (JSON) generated by the back-end, and dynamically creates DOM elements in combination with a preset field component library (such as UI components encapsulated with text boxes, checkboxes, date pickers, etc.).

[0076] Bind interactive logic and verification rules:

[0077] Based on the events (such as onChange) and validation rules (such as required:true) defined in the template, the front-end binds event handlers (such as input validation and field linkage) via JavaScript to ensure that form interactions conform to custom logic. For example, when "role" is selected as "administrator", an additional "permission settings" field is displayed using {@if role === 'admin'}.

[0078] Example 2:

[0079] Combined with appendix Figure 2 As shown, a production safety inspection method in a classified offline environment includes:

[0080] Basic information (i.e., basic data) is stored by entering QR codes or graphic locks. The basic information includes location information and personnel information.

[0081] The old form data is divided into specific inspection items, and the inspection items are categorized into inspection items to obtain a custom form;

[0082] Establish and store the association between basic information and custom forms for quick retrieval of the corresponding information;

[0083] Based on the basic information and relationships, the corresponding custom forms are retrieved and checked, information is corrected and filled in, and result data is generated, including structured test results or QR codes for single results, thereby realizing the structuring of test data and improving the efficiency of check form filling.

[0084] The test results can be transmitted to a standalone or isolated environment via USB or by scanning a QR code.

[0085] Furthermore, the process of segmenting the old form data into specific inspection content and categorizing the inspection content into inspection items to obtain a custom form can be broken down into four core steps: "old data parsing → content segmentation → rule matching → categorized storage".

[0086] 1. Data Model Design (Prerequisite):

[0087] The core data structure needs to be defined in advance (taking a relational database as an example):

[0088] The old form data table old_form_data stores the original form data. The fields must include at least the unique form identifier form_id, the original form content raw_content (which may be JSON / text / rich text), and the form type form_type (used to distinguish the parsing rules).

[0089] The check_item table stores standardized check items, with fields such as check item ID (item_id), check content (e.g., "Is the door closed?"), and check method (method).

[0090] The `check_content` table stores the specific content extracted from the old data, with fields such as `content_id`, associated old form `form_id`, associated check item ID `item_id()`, specific value `content_value` (e.g., "closed and locked"), and additional information `ext_info` (e.g., unit, remarks).

[0091] 2. Parsing old form data (extracting original content)

[0092] Read old data: Query the old form data table old_form_data through an interface (such as getOldFormData(formId)) to obtain the original form content raw_content and form type form_type of the target form.

[0093] Structured parsing: Select the corresponding parsing strategy based on the form type (form_type) (avoid hardcoding; this can be achieved through the strategy pattern): If the raw form content (raw_content) is JSON (e.g., {"Check content": door is closed", "Remarks": "None"}), parse it directly into a JSON object and extract key fields; if it is text / rich text (e.g., "1. door is closed; 2. power is off"), extract key-value pairs (parsedData) using regular expressions (e.g., matching "door: (.*?);"); if it is tabular data (e.g., HTML table), parse the tags and map them to key-value pairs (parsedData) by row and column.

[0094] 3. Check content segmentation (break down into specific items)

[0095] Segmentation by Rules: Based on the parsed key-value pairs `parsedData`, the original content is split into the smallest granularity of "inspection content" (i.e., indivisible key-value pairs). For example, the original content "Door is closed" is split into: key: "Door", value: "Closed". Handling complex scenarios: If nested content exists (such as "Safety inspection: [Fire equipment: intact; Passage: unobstructed]"), the nested structure is split through recursive parsing; if there are ambiguous expressions (such as "All indicators meet the standards"), they are mapped to preset inspection content based on business rules (such as associating default values ​​such as "Indicator A" and "Indicator B").

[0096] 4. Categorization of inspection items (matching and storage)

[0097] Establish a matching rule base: Pre-define the mapping relationship between "check content key name" and "check item check_item", which can be stored in a database or configuration file, such as Table 1:

[0098] Table 1. Mapping Relationship between Inspection Content Key Names and Inspection Items

[0099]

[0100] Perform categorization matching: Iterate through the split key-value pairs parsedData; for each key name, match the corresponding item_id using the rule base (supports exact matching, fuzzy matching, and synonym mapping, which can be combined with simple NLP to improve matching accuracy); if the match is successful, generate a CheckContent object and associate it with form_id, item_id, and content_value; if the match fails, mark it as "uncategorized content" and trigger manual intervention (such as prompting the user to manually associate the check item in the management interface).

[0101] 5. Results Feedback and Iteration

[0102] After classification, the processing results are returned via an interface (such as getClassificationResult(formId)), including successfully classified check items and the number of unclassified items. Manual correction of unclassified items is supported (e.g., manually selecting the corresponding check item in the front-end interface). After correction, the check_content table is automatically updated, and the new mapping relationship is synchronized to the rule base, improving the accuracy of subsequent automatic matching. Alternatively, check items and detection content can be added through the front-end interface, and then a new check form can be constructed by selecting the required check items and content through the form management module.

[0103] Based on the aforementioned basic information and relationships, the corresponding custom form is retrieved, checked, corrected, and populated to generate result data. This process includes form retrieval → basic information population → structured data entry → result generation and QR code output.

[0104] I. Form Template Caching: Each form has a corresponding original template, pre-set and stored in the database. Quick retrieval based on specific scenarios:

[0105] Triggering methods: The front end provides three entry points: "Recently Used", "Filter by Type", and "Scan to Retrieve" (scan the device / scene QR code to obtain the associated form_id);

[0106] II. Automatic filling of basic information (reducing repetitive data entry):

[0107] Basic information data source:

[0108] 1) Pre-stored data: The backend provides the current user information (name, employee number, etc.) through the interface, and obtains the inspection item information (such as inspection content, inspection method, inspection source, inspection result, etc.) and stores it in the frontend store (such as Vuex).

[0109] 2) Associated data: If the form is associated with a specific scenario (such as "Workshop A-Section 001"), the scene_id is carried when the form is retrieved, and the backend returns the preset basic data of that scenario (such as the inspection location and the person in charge).

[0110] 3) For some common information, such as entering an employee's ID number, the name and job title will be automatically populated.

[0111] III. Structured Data Entry for Testing (Standardized Data Format):

[0112] 1) Formatting: Automatically format fields such as dates and numbers (e.g., inputting "35" will automatically complete to "35℃", and dates will be converted to "YYYY-MM-DD HH:mm").

[0113] 2) Structured data storage:

[0114] 3) Front-end assembly: After the user completes the form, the front-end converts the formData (key-value pairs) into structured JSON, containing form_id, basic_info (a subset of basic information), detection_data (a subset of detection data), submit_time, etc.

[0115] 4) Backend storage: Submit data through the / api / detection / save interface, and the backend stores it in the detection_result table (structured fields), while retaining the original JSON string (raw_data field) for traceability.

[0116] IV. Result Generation and QR Code Output:

[0117] 1) Generation of structured detection results

[0118] Backend Interface: Provides an interface that returns structured results (supports JSON / Excel format). The Excel file is dynamically generated using the POI library (rows and columns are organized according to form fields).

[0119] Front-end display: Parse the structured data returned by the interface and display it using table or card components, supporting file download triggered by the "Export" button.

[0120] 2) Single result QR code generation

[0121] Data encoding: The backend generates a unique result_id for each test result, and the QR code content is the URL of the result details page or structured data JSON (which needs to be compressed).

[0122] QR code generation: Front-end generation: Using the qrcode.js library, a QR code image is generated directly in the browser based on result_id or structured data. In addition, the QR code can be encrypted before transmission.

[0123] This invention, through the classification and design of security inspection elements and the collaboration of various functional modules, satisfies the diverse customization needs of forms under the premise of data structuring. Under the requirement of wireless network isolation, it also meets the need for structured data transmission (this mechanism increases the data transmission success rate in wireless network isolation environments from 72% to 99.6% of traditional solutions, and controls the transmission time of a single inspection data entry to within 100ms, while eliminating formatting errors caused by unstructured data transmission and significantly reducing the workload of backend data cleaning (cleaning time reduced by 80%)). Furthermore, the rapid form filling improves the efficiency of on-site inspections, ultimately providing better assurance for the accumulation of basic data for security inspections.

[0124] Although the present invention has been described herein with reference to illustrative embodiments, the above embodiments are merely preferred embodiments of the present invention, and the implementation of the present invention is not limited to the above embodiments. It should be understood that those skilled in the art can devise many other modifications and implementations, which will fall within the scope and spirit of the principles disclosed in this application.

Claims

1. A production safety inspection system for a classified offline environment, characterized in that, include: The basic information module is configured to store basic information, including location information and personnel information. The form management module is configured to split old form data into specific inspection contents and categorize the inspection contents into inspection items to obtain custom forms; The association module is configured to establish and store the association between the basic information and the custom form; The form filling module is configured to retrieve the corresponding custom form based on the basic information and relationships, and perform checks, information corrections and information filling to generate result data. The result transmission module is configured to transmit the result data offline.

2. The production safety inspection system for a classified offline environment according to claim 1, characterized in that, The form management module includes: The data model design unit is configured to define the data structures of the old form data table old_form_data, the check item table check_item, and the check content table check_content. The old data parsing unit is configured to retrieve the original form content and form type of the target form by querying the old form data table; and to select the corresponding parsing strategy according to the form type to perform structured parsing to obtain key fields or key-value pairs parsedData. The content segmentation unit is configured to split the original content into the smallest granularity of inspection content based on the parsed key-value pairs parsedData, or to map it to preset inspection content in combination with business rules. The classification and matching unit is configured to establish a matching rule base. The matching rule base has a preset mapping relationship between the key name of the check content and the check item. It matches the key-value pair parsedData with the corresponding check item ID item_id. If the match is successful, a CheckContent object is generated. The CheckContent object is associated with the unique form ID form_id, the check item ID item_id, and the specific value content_value. Otherwise, manual intervention is triggered. The unit also provides feedback on the processing results and automatically updates the check content table check_content after manual correction, and synchronizes the new mapping relationship to the matching rule base. The front-end interface is configured to receive and display the processing results, allowing users to manually select corresponding inspection items, correct unclassified content, and add inspection items and content, and then feed back to the classification and matching unit.

3. The production safety inspection system for a classified offline environment according to claim 2, characterized in that, The old form data table old_form_data is used to store the original form data. The fields include the unique form identifier form_id, the original form content raw_content, and the form type form_type.

4. The production safety inspection system for a classified offline environment according to claim 2, characterized in that, The check_item table is used to store standardized check items, and its fields include check item ID, check content, and check method.

5. A production safety inspection system for a classified offline environment according to claim 3, characterized in that, The check_content table is used to store the specific content extracted from the original form data. The fields include content ID content_id, associated form unique identifier form_id, associated check item ID item_id, specific value content_value, and additional information ext_info.

6. A production safety inspection system for a classified offline environment according to claim 2, characterized in that, The processing results include successfully categorized inspection content, the number of uncategorized content, and manually corrected uncategorized content.

7. A production safety inspection system for a classified offline environment according to claim 1, characterized in that, The result data is a structured detection result or a QR code of a single result. The result data is transmitted offline via USB or by scanning a QR code to a standalone or isolated environment.

8. A production safety inspection system for a classified offline environment according to claim 1, characterized in that, It also includes a QR code recognition module and a graphic lock module. The QR code recognition module is configured to establish a unique association between the location and the QR code by encoding the location information into the QR code and storing it in the basic information module. The graphic lock module is configured to establish a unique association between the graphic lock and personnel information by storing the information drawn by the graphic lock in a string and storing it in the basic information module.

9. A method for production safety inspection in a classified offline environment, characterized in that, include: Store basic information, including location information and personnel information; The old form data is divided into specific inspection items, and the inspection items are categorized into inspection items to obtain a custom form; Establish and store the association between the basic information and the custom form; Based on the basic information and relationships, retrieve the corresponding custom form and perform checks, information corrections, and information filling to generate result data; The resulting data will be transmitted offline.

10. A production safety inspection method in a classified offline environment according to claim 9, characterized in that, The process of dividing the old form data into specific inspection contents and categorizing the inspection contents into inspection items to obtain a custom form specifically includes: Define the data structures for the old form data table `old_form_data`, the check item table `check_item`, and the check content table `check_content`. The old form data table `old_form_data` stores the original form data, with fields including the unique form identifier `form_id`, the original form content `raw_content`, and the form type `form_type`. The check item table `check_item` stores standardized check items, with fields including the check item ID `item_id`, the check content `content`, and the check method `method`. The check content table `check_content` stores the specific content extracted from the original form data, with fields including the content ID `content_id`, the associated unique form identifier `form_id`, the associated check item ID `item_id`, the specific value `content_value`, and additional information `ext_info`. By querying the old form data table, the original form content and form type of the target form can be obtained; and the corresponding parsing strategy can be selected according to the form type to perform structured parsing to obtain key fields or key-value pairs parsedData; Based on the parsed key-value pairs, the original content is split into the smallest granularity of inspection content, or mapped to preset inspection content in combination with business rules; Based on the pre-defined mapping relationship between check content keys and check items in the matching rule base, the key-value pair `parsedData` is matched against the corresponding check item ID `item_id`. If the match is successful, a `CheckContent` object is generated, which is associated with the unique form ID `form_id`, the check item ID `item_id`, and the specific value `content_value`. Otherwise, manual intervention is triggered, and the processing results are fed back, including the successfully categorized check content, the number of uncategorized content, and the manual correction of uncategorized content. The system receives and displays the processing results, allowing users to manually select corresponding check items, correct unclassified content, and add new check items and content. After manual correction, the system automatically updates the check content table (check_content) and synchronizes the new mapping relationship to the matching rule base.

Images