Bridge engineering quantity management method and device based on multi-end screenshot and WeChat push

Abstract

The present application relates to a kind of bridge engineering quantity management methods based on multi-terminal screenshot and WeChat push, comprising: multi-terminal image acquisition and interactive frame selection;Multi-modal large model identification and data mapping;Multi-dimensional search and historical traceability;Built-in parameter automatic completion;WeChat regular push;Format report output.The present application significantly improves the convenience of data collection;Fundamentally solve the problem that the recognition result is only a stream of unstructured text, still needs manual secondary input and checking;Solve the problem that data is stored in independent Excel file under traditional mode, the search capability is weak, realize efficient search according to component space level, material specification keyword, multi-field combination condition;Provide complete change track for design change, site certification and other data adjustment in construction process, meet the compliance requirements of engineering audit and dispute traceability;Automatic message and attachment are sent to specified group, ensure that material release notice is timely and accurately delivered.

Description

Technical Field

[0001] This invention relates to the field of bridge engineering quantity management technology, and in particular to a bridge engineering quantity management method and device based on multi-terminal screenshots and WeChat push notifications. Background Technology

[0002] In bridge construction management, the quantity data in design drawings, material lists, and change orders are the core basis for material issuance, progress tracking, measurement and payment, and cost accounting. Existing technologies suffer from the following significant drawbacks:

[0003] First, the data collection entry point is singular and lacks sufficient scenario coverage: existing tools either only support uploading photos taken with a mobile phone or only support importing local files from a computer. When technicians view electronic drawings in the office, the most convenient way is to use a screenshot function similar to WeChat's Alt+A to directly select the table area, rather than opening a dedicated app to take a photo. The lack of a unified data collection entry point that supports both quick screenshots on PCs and photos taken on mobile devices leads to fragmented operations and low efficiency in the data collection process;

[0004] Second, manual secondary input is still required after collection. Traditional OCR lacks semantic understanding: general OCR tools have limited accuracy in recognizing engineering charts containing complex table lines and merged cells, and the output after recognition is an unstructured text stream. There is a lack of a mechanism that combines multimodal large models for semantic-level recognition and interactive correction.

[0005] Third, data storage is scattered and retrieval capabilities are weak: even if data is obtained through OCR or manual input, this data is often stored in the form of independent Excel files on personal computers or WeChat groups, lacking a relational database management system with components as the core and supporting multi-dimensional combined retrieval.

[0006] Fourth, historical data changes are untraceable, making auditing difficult: During construction, design changes, on-site approvals, and other reasons often lead to adjustments in engineering quantity data. Once a data dispute occurs or an audit is scheduled, it is impossible to provide a complete change tracker.

[0007] Fifth, the methods for notifying material requisition information are rudimentary and lack timeliness: After the material requisition form is completed, it is usually photographed and sent to the WeChat work group by the materials department staff. Relevant personnel need to check the group messages themselves, which is prone to omissions or delays. There is a lack of a timed, automatic, and template-based group notification mechanism.

[0008] In summary, there is an urgent need for an integrated bridge engineering quantity data management method that integrates PC / mobile dual-terminal screenshot collection, multimodal large model semantic recognition, relational database multidimensional retrieval and audit traceability, WeChat scheduled push, and built-in parameter automatic completion. Summary of the Invention

[0009] To address the problems of existing technologies, such as a single data acquisition entry point, reliance on general-purpose OCR tools for data recognition, fragmented data management, and weak application capabilities, the primary objective of this invention is to provide a bridge engineering quantity management method based on multi-terminal screenshots and WeChat push notifications. This method enables a unified dual-terminal acquisition entry point that combines quick screenshots on PCs and photos taken on mobile devices. It also allows for text recognition and semantic analysis of selected areas in the image, and enables efficient retrieval based on component spatial hierarchy, material specification keywords, and multi-field combination conditions.

[0010] To achieve the above objectives, the present invention adopts the following technical solution: a bridge engineering quantity management method based on multi-terminal screenshots and WeChat push notifications, the method comprising the following sequential steps:

[0011] (1) Multi-terminal image acquisition and interactive box selection: acquire engineering chart images, generate at least one rectangular selection box through the area selection operation triggered by the user in the image preview interface, and specify a data type from the preset label library for each rectangular selection box;

[0012] (2) Multimodal large model recognition and data mapping: The images of each rectangular selection area are cropped into slices, and the slices and corresponding labels are sent to the multimodal large model. The multimodal large model performs text recognition and semantic parsing, outputs the recognition results, and maps the recognition results into structured data records according to the labels. After user confirmation, the data is stored in a relational database and written to the operation audit log.

[0013] (3) Multidimensional retrieval and historical traceability: For the main table of engineering quantity details, the component structure tree table and the extended table in the relational database, an inverted index structure of text fields is established, and query cache units and materialized views are constructed to obtain an enhanced relational database that supports multidimensional combined retrieval and efficient aggregate query; According to the search conditions input by the user, the search results are comprehensively sorted and matched by using the inverted index, query cache and AI-assisted sorting, and the sorted engineering quantity data records are returned; According to the historical traceability request, the change records matched by the operation audit log table are queried, and the original value before modification, the new value after modification and the operator information are displayed in chronological order;

[0014] (4) Built-in parameter auto-completion: The steel diameter field stored in the main table of engineering quantity details and the extended table of steel reinforcement details in the enhanced relational database is retrieved, the built-in theoretical weight parameter table of steel reinforcement is called, the corresponding theoretical weight per linear meter is automatically matched according to the steel diameter, the unit weight field and the total weight field are calculated and backfilled, and the completed engineering quantity details database is obtained.

[0015] (5) Scheduled WeChat push: Through the WeChat interface, according to the preset push time and message template, the material issuance data is retrieved from the completed engineering quantity details database at regular intervals, and the message and attachment are generated by the template engine and pushed.

[0016] (6) Formatted report output: Based on the template selected by the user, aggregate data from the completed engineering quantity details database to generate and export formatted business reports.

[0017] In step (2), the multimodal large model adopts the Qwen-VL, GPT-4V, or Claude Vision visual models; the multimodal large model receives the selected area image and label type as prompt words, and outputs structured JSON data containing the recognized text and confidence level; the relational database includes a project table, a component structure tree table, a main table of engineering quantity details, an extended table of steel reinforcement details, an extended table of concrete details, an operation audit log table, a WeChat push log table, and a material overrun analysis report; among them, the component structure tree table is associated with the project table through project identifiers and self-associations through parent node identifiers to achieve a hierarchical structure; the main table of engineering quantity details is associated with the component structure tree table through component identifiers and with the project table through project identifiers; the extended table of steel reinforcement details and the extended table of concrete details are associated with the main table of engineering quantity details through detail identifiers; the operation audit log table records all add, delete, and modify operations of all business tables.

[0018] In step (3), establishing the inverted index structure for the text field and constructing the query cache unit and materialized view specifically includes the following steps:

[0019] (3a) Create an inverted index for the text fields of specifications, material names, and component names in the main table of quantities to accelerate the response speed of keyword retrieval;

[0020] (3b) The results of high-frequency aggregation queries are cached, and the cache hit rate H satisfies the following relationship:

[0021] ;

[0022] in, To cache the hit count, To cache the number of cache misses;

[0023] (3c) Materialized views are pre-created for statistical requirements of grouping and summarizing. The materialized views are pre-aggregated according to the hierarchical relationship of the component structure tree table and the specification field of the engineering quantity details main table to accelerate the response speed of grouping and summarizing queries.

[0024] In step (3), the AI-assisted sorting uses the following formula to comprehensively sort the search results:

[0025] ;

[0026] In the formula: For document The overall ranking score; For query terms With Documents BM25 correlation score; The current time; For document The creation or modification time; It is the attenuation constant; For document The number of times the historical data has been searched and clicked; All are configurable weight coefficients, satisfying .

[0027] In step (4), the theoretical weight parameter table for reinforcing bars includes the weight per linear meter for commonly used specifications with diameters from 6mm to 32mm; when a material requisition form is generated, the unit weight is automatically obtained from the table based on the diameter of the reinforcing bars, and the total weight is calculated. :

[0028] ;

[0029] in, For single length, For roots, denoted as the theoretical weight per linear meter, and n represents the total number of steel reinforcement detail records.

[0030] In step (5), the WeChat scheduled push includes: setting the daily push time point, target group identifier, and message template ID; scheduling the scheduled task to trigger data retrieval according to the configured time; combining the retrieved material delivery data with the template engine to generate a text message, and calling the Excel generation library to create a material delivery form attachment; sending the message through the WeChat Webhook API, and recording the push status and results in the WeChat push log table.

[0031] In step (6), when the user selects the material quantity deviation report template, the engineering quantity deviation rate is automatically calculated. :

[0032] ;

[0033] In the formula, This represents the actual amount of work done. For the design engineering quantity;

[0034] when When this happens, it will be highlighted in the report using a preset marker.

[0035] Another object of the present invention is to provide an electronic device comprising:

[0036] Processor; and

[0037] The memory stores computer program instructions, which, when executed by the processor, cause the processor to perform the bridge engineering quantity management method based on multi-terminal screenshots and WeChat push, as described above.

[0038] The present invention also provides a computer-readable storage medium having stored thereon computer program instructions, which, when executed by a processor, cause the processor to perform the bridge engineering quantity management method based on multi-terminal screenshots and WeChat push as described above.

[0039] As can be seen from the above technical solution, the beneficial effects of the present invention are as follows: First, the present invention realizes a unified data acquisition entry point for both PC-based quick screenshots and mobile-based photo capture. By generating a rectangular selection box in response to user-triggered region selection operations and specifying the data type for the selection box, image cropping and data classification and labeling are integrated into a single operation. This solves the problem of fragmented operation and low efficiency caused by using screenshot tools and OCR tools separately in the prior art, significantly improving the convenience of data acquisition. Second, the present invention introduces a multimodal large model to perform text recognition and semantic analysis on the selected region image. Compared with traditional OCR tools, it can understand text containing complex table lines. First, the invention merges the engineering chart features of cells and, combined with user-specified label types, directly maps the recognition results into structured data records, outputting structured JSON data containing the recognized text and confidence levels. After user confirmation, this data is stored in the database, fundamentally solving the problem that the recognition results are merely unstructured text streams requiring manual secondary entry and verification. Second, the invention constructs a data management system centered on a relational database. This system includes a project table, a component structure tree table, a detailed engineering quantity main table, and extended tables. Based on this, an inverted index structure, query cache units, and materialized views are built, resulting in a system that supports multi-dimensional combined retrieval and high performance. This invention provides an enhanced relational database for efficient aggregation and querying, solving the problems of traditional methods where data is stored in scattered, independent Excel files and has weak retrieval capabilities. It achieves efficient retrieval based on component spatial hierarchy, material specification keywords, and multi-field combinations. Fourth, this invention completely records all add, delete, and modify operations in all business tables through an operation audit log table. Responding to historical traceability requests, it can display data values ​​before and after changes and operator information in chronological order, providing a complete change trajectory for design changes, on-site approvals, and other data adjustments during construction, meeting the compliance requirements for engineering audits and dispute tracing. Fifth, this invention incorporates the theoretical weight of reinforcing steel bars. The parameter table automatically retrieves the theoretical weight per linear meter corresponding to the diameter of the rebar when generating documents, calculates and fills in the unit weight and total weight fields, and obtains a complete engineering quantity details database, avoiding errors that may be introduced by manual table lookup and calculation, and ensuring the accuracy and consistency of the data; Sixth, this invention realizes timed push through the WeChat interface, changing "people looking for information" to "information finding people". According to the preset push time and message template, it retrieves material issuance data from the complete engineering quantity details database at regular intervals, automatically generates messages and attachments and sends them to designated groups, ensuring that material issuance notices are delivered in a timely and accurate manner, and solving the problem of omissions or delays caused by existing manual photo group sending. Attached Figure Description

[0040] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0041] like Figure 1As shown, a method for managing bridge engineering quantities based on multi-terminal screenshots and WeChat push notifications is presented. This method includes the following steps in sequence:

[0042] (1) Multi-terminal image acquisition and interactive box selection: acquire engineering chart images, generate at least one rectangular selection box through the area selection operation triggered by the user in the image preview interface, and specify a data type from the preset label library for each rectangular selection box;

[0043] (2) Multimodal large model recognition and data mapping: The images of each rectangular selection area are cropped into slices, and the slices and corresponding labels are sent to the multimodal large model. The multimodal large model performs text recognition and semantic parsing, outputs the recognition results, and maps the recognition results into structured data records according to the labels. After user confirmation, the data is stored in a relational database and written to the operation audit log.

[0044] (3) Multidimensional retrieval and historical traceability: For the main table of engineering quantity details, the component structure tree table and the extended table in the relational database, an inverted index structure of text fields is established, and query cache units and materialized views are constructed to obtain an enhanced relational database that supports multidimensional combined retrieval and efficient aggregate query; According to the search conditions input by the user, the search results are comprehensively sorted and matched by using the inverted index, query cache and AI-assisted sorting, and the sorted engineering quantity data records are returned; According to the historical traceability request, the change records matched by the operation audit log table are queried, and the original value before modification, the new value after modification and the operator information are displayed in chronological order;

[0045] (4) Built-in parameter auto-completion: The steel diameter field stored in the main table of engineering quantity details and the extended table of steel reinforcement details in the enhanced relational database is retrieved, the built-in theoretical weight parameter table of steel reinforcement is called, the corresponding theoretical weight per linear meter is automatically matched according to the steel diameter, the unit weight field and the total weight field are calculated and backfilled, and the completed engineering quantity details database is obtained.

[0046] (5) Scheduled WeChat push: Through the WeChat interface, according to the preset push time and message template, the material issuance data is retrieved from the completed engineering quantity details database at regular intervals, and the message and attachment are generated by the template engine and pushed.

[0047] (6) Formatted report output: Based on the template selected by the user, aggregate data from the completed engineering quantity details database to generate and export formatted business reports.

[0048] In step (2), the multimodal large model adopts the Qwen-VL, GPT-4V, or Claude Vision visual models; the multimodal large model receives the selected area image and label type as prompt words, and outputs structured JSON data containing the recognized text and confidence level; the relational database includes a project table, a component structure tree table, a main table of engineering quantity details, an extended table of steel reinforcement details, an extended table of concrete details, an operation audit log table, a WeChat push log table, and a material overrun analysis report; among them, the component structure tree table is associated with the project table through project identifiers and self-associations through parent node identifiers to achieve a hierarchical structure; the main table of engineering quantity details is associated with the component structure tree table through component identifiers and with the project table through project identifiers; the extended table of steel reinforcement details and the extended table of concrete details are associated with the main table of engineering quantity details through detail identifiers; the operation audit log table records all add, delete, and modify operations of all business tables.

[0049] In step (3), establishing the inverted index structure for the text field and constructing the query cache unit and materialized view specifically includes the following steps:

[0050] (3a) Create an inverted index for the text fields of specifications, material names, and component names in the main table of quantities to accelerate the response speed of keyword retrieval;

[0051] (3b) The results of high-frequency aggregation queries are cached, and the cache hit rate H satisfies the following relationship:

[0052] ;

[0053] in, To cache the hit count, To cache the number of cache misses;

[0054] (3c) Materialized views are pre-created for statistical requirements of grouping and summarizing. The materialized views are pre-aggregated according to the hierarchical relationship of the component structure tree table and the specification field of the engineering quantity details main table to accelerate the response speed of grouping and summarizing queries.

[0055] In step (3), the AI-assisted sorting uses the following formula to comprehensively sort the search results:

[0056] ;

[0057] In the formula: For document The overall ranking score; For query terms With Documents BM25 correlation score; The current time; For document The creation or modification time; It is the attenuation constant; For document The number of times the historical data has been searched and clicked; All are configurable weight coefficients, satisfying .

[0058] In step (4), the theoretical weight parameter table for reinforcing bars includes the weight per linear meter for commonly used specifications with diameters from 6mm to 32mm; when a material requisition form is generated, the unit weight is automatically obtained from the table based on the diameter of the reinforcing bars, and the total weight is calculated. :

[0059] ;

[0060] in, For single length, For roots, denoted as the theoretical weight per linear meter, and n represents the total number of steel reinforcement detail records.

[0061] In step (5), the WeChat scheduled push includes: setting the daily push time point, target group identifier, and message template ID; scheduling the scheduled task to trigger data retrieval according to the configured time; combining the retrieved material delivery data with the template engine to generate a text message, and calling the Excel generation library to create a material delivery form attachment; sending the message through the WeChat Webhook API, and recording the push status and results in the WeChat push log table.

[0062] In step (6), when the user selects the material quantity deviation report template, the engineering quantity deviation rate is automatically calculated. :

[0063] ;

[0064] In the formula, This represents the actual amount of work done. For the design engineering quantity;

[0065] when When this happens, it will be highlighted in the report using a preset marker.

[0066] Example 1

[0067] Material issuance management application for the [XX Bridge] project of a certain expressway

[0068] First, initialize and configure the desktop application by logging into the desktop management system, creating a project "[XX Expressway XX Section]", filling in the project number and other information, and writing it into the project information table. Import components in batches from Excel or manually create a component structure tree and store it in the component structure tree table, forming a hierarchy of "XX Bridge → A Ramp → 1# Pier → Pile Foundation…". Go to "System Settings - Parameter Management" and confirm the theoretical weight table data for the reinforcing steel. Configure the WeChat Webhook address and set the push rules: "Push to the reinforcing steel processing group at 8:30 AM daily" and "Push to the site management group at 4:00 PM daily". Drag and drop fields in the template designer to define the material issuance column and grouping.

[0069] Next, image acquisition, selection, recognition, and data entry are performed: Office scenario: Open the PDF drawing, press Ctrl+Shift+A to enter screenshot mode, and sequentially select the columns "Rebar Number," "Specification," "Number of Rebars," and "Single Length," assigning labels to each. Click "Complete and Recognize," and the system sends the four selected slices and labels to the Qwen-VL multimodal large model, returning JSON data. After verification, the component "A Ramp → 3# Pier → Cap Beam" is selected and entered into the database. The data is written to the main table of engineering quantity details and the extended table of rebar details, and the operation audit log is recorded. On-site scenario: Take a picture of the hand-modified notification form, select "Station Number," "Extended Main Rebar Length," and "Main Rebar Specification" in the APP, note the reason for the change, upload it, synchronize it to the cloud, and leave a record.

[0070] Next, multi-dimensional retrieval and tracing are performed: After reviewing the data, enter "Φ25" in the search box to execute a specification search. The system uses an inverted index and cache to quickly return the components and total quantity that use Φ25 in the entire bridge. Clicking on the results will locate the component tree, and clicking on "Change History" will display the change records and original screenshots, making the audit trail clear.

[0071] Next, the material requisition form is generated and automatically pushed to WeChat: The items "3# Pier Cap Beam" and "4# Pier Pile Foundation" are selected, and the system automatically looks up the table to complete the unit weight and calculates the total weight (e.g., G=27.6m×24×3.85≈2549.8kg). The Apache POI document generation library is then used to generate an Excel material requisition form. At 8:30 AM the following morning, a scheduled task is triggered, retrieving approved but not yet pushed records, rendering the message using the FreeMarker template engine, uploading attachments, and pushing it to the rebar processing group via WeChat, recording the push log.

[0072] Finally, the report is output:

[0073] At the end of the month, select a time period and export a summary table of material issuance grouped by specifications, as well as a material saving / overrun analysis table. By tracing the complete change history, the audit can be completed efficiently.

[0074] In summary, this invention achieves a unified data acquisition entry point for both PC-based quick screenshots and mobile-based photo capture. By generating a rectangular selection box in response to user-triggered region selection operations and assigning data type to the selection box, it integrates image cropping and data classification and labeling into a single operation. This solves the problem of fragmented operation and low efficiency caused by using separate screenshot tools and OCR tools in existing technologies, significantly improving the convenience of data acquisition. Furthermore, this invention introduces a multimodal large model to perform text recognition and semantic analysis on the selected region image. Compared to traditional OCR tools, it can understand the features of engineering charts containing complex table lines and merged cells. This invention, combined with user-specified tag types, directly maps the recognition results into structured data records, outputting structured JSON data containing the recognized text and confidence levels. After user confirmation, this data is stored in the database, fundamentally solving the problem that the recognition results are merely unstructured text streams requiring manual secondary entry and verification. The invention constructs a data management system centered on a relational database. By establishing a project table, component structure tree table, engineering quantity details master table, and extended tables, and on this basis, building an inverted index structure, query cache units, and materialized views, it obtains an enhanced relational database supporting multi-dimensional combined retrieval and efficient aggregation queries. This solves the problem of traditional methods where data is scattered and stored in independent Excel files, resulting in weak retrieval capabilities. It achieves efficient retrieval by component spatial hierarchy, material specification keywords, and multi-field combination conditions. The invention completely records all addition, deletion, and modification operations of business tables through an operation audit log table. Responding to historical traceability requests, it can display the data before and after changes and operator information in chronological order, providing a complete change trajectory for design changes, on-site visas, and other data adjustments during construction, meeting the compliance requirements of engineering audits and dispute tracing. The invention also incorporates a built-in theoretical weight parameter table for steel reinforcement. When generating a document, the system automatically retrieves the theoretical weight per linear meter corresponding to the diameter of the reinforcing bar, calculates and fills in the unit weight and total weight fields, and obtains a complete engineering quantity details database. This avoids errors that may be introduced by manual table lookup and calculation, ensuring the accuracy and consistency of the data. The invention also enables timed push notifications through the WeChat interface, changing "people searching for information" to "information finding people." Based on the preset push time and message template, the system periodically retrieves material issuance data from the complete engineering quantity details database, automatically generates messages and attachments, and sends them to designated groups. This ensures that material issuance notifications are delivered in a timely and accurate manner, solving the problem of omissions or delays caused by manual photo-taking and mass sending in the past.

[0075] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A method for managing bridge engineering quantities based on multi-terminal screenshots and WeChat push notifications, characterized in that: The method includes the following steps in sequence: (1) Multi-terminal image acquisition and interactive box selection: acquire engineering chart images, generate at least one rectangular selection box through the area selection operation triggered by the user in the image preview interface, and specify a data type from the preset label library for each rectangular selection box; (2) Multimodal large model recognition and data mapping: The images of each rectangular selection area are cropped into slices, and the slices and corresponding labels are sent to the multimodal large model. The multimodal large model performs text recognition and semantic parsing, outputs the recognition results, and maps the recognition results into structured data records according to the labels. After user confirmation, the data is stored in a relational database and written to the operation audit log. (3) Multidimensional retrieval and historical traceability: For the main table of engineering quantity details, the component structure tree table and the extended table in the relational database, an inverted index structure of text fields is established, and query cache units and materialized views are constructed to obtain an enhanced relational database that supports multidimensional combined retrieval and efficient aggregate query; According to the search conditions input by the user, the search results are comprehensively sorted and matched by using the inverted index, query cache and AI-assisted sorting, and the sorted engineering quantity data records are returned; According to the historical traceability request, the change records matched by the operation audit log table are queried, and the original value before modification, the new value after modification and the operator information are displayed in chronological order; (4) Built-in parameter auto-completion: The steel diameter field stored in the main table of engineering quantity details and the extended table of steel reinforcement details in the enhanced relational database is retrieved, the built-in theoretical weight parameter table of steel reinforcement is called, the corresponding theoretical weight per linear meter is automatically matched according to the steel diameter, the unit weight field and the total weight field are calculated and backfilled, and the completed engineering quantity details database is obtained. (5) Scheduled WeChat push: Through the WeChat interface, according to the preset push time and message template, the material issuance data is retrieved from the completed engineering quantity details database at regular intervals, and the message and attachment are generated by the template engine and pushed. (6) Formatted report output: Based on the template selected by the user, aggregate data from the completed engineering quantity details database to generate and export formatted business reports.

2. The bridge engineering quantity management method based on multi-terminal screenshots and WeChat push according to claim 1, characterized in that: In step (2), the multimodal large model adopts the Qwen-VL, GPT-4V, or ClaudeVision visual model; the multimodal large model receives the selected area image and label type as prompt words, and outputs structured JSON data containing the recognized text and confidence level; the relational database includes a project table, a component structure tree table, a main table of engineering quantity details, an extended table of steel reinforcement details, an extended table of concrete details, an operation audit log table, a WeChat push log table, and a material overrun analysis report; among them, the component structure tree table is associated with the project table through project identifiers and self-associations through parent node identifiers to achieve a hierarchical structure; the main table of engineering quantity details is associated with the component structure tree table through component identifiers and with the project table through project identifiers; the extended table of steel reinforcement details and the extended table of concrete details are associated with the main table of engineering quantity details through detail identifiers; the operation audit log table records all add, delete, and modify operations of all business tables.

3. The bridge engineering quantity management method based on multi-terminal screenshots and WeChat push according to claim 1, characterized in that: In step (3), establishing the inverted index structure for the text field and constructing the query cache unit and materialized view specifically includes the following steps: (3a) Create an inverted index for the text fields of specifications, material names, and component names in the main table of quantities to accelerate the response speed of keyword retrieval; (3b) The results of high-frequency aggregation queries are cached, and the cache hit rate H satisfies the following relationship: ； in, To cache the hit count, To cache the number of cache misses; (3c) Materialized views are pre-created for statistical requirements of grouping and summarizing. The materialized views are pre-aggregated according to the hierarchical relationship of the component structure tree table and the specification field of the engineering quantity details main table to accelerate the response speed of grouping and summarizing queries.

4. The bridge engineering quantity management method based on multi-terminal screenshots and WeChat push according to claim 1, characterized in that: In step (3), the AI-assisted sorting uses the following formula to comprehensively sort the search results: ； In the formula: For document The overall ranking score; For query terms With Documents BM25 correlation score; The current time; For document The creation or modification time; It is the attenuation constant; For document The number of times the historical data has been searched and clicked; All are configurable weight coefficients, satisfying .

5. The bridge engineering quantity management method based on multi-terminal screenshots and WeChat push according to claim 1, characterized in that: In step (4), the theoretical weight parameter table for reinforcing bars includes the weight per linear meter for commonly used specifications with diameters from 6mm to 32mm; when a material requisition form is generated, the unit weight is automatically obtained from the table based on the diameter of the reinforcing bars, and the total weight is calculated. : ； in, For single length, For roots, denoted as the theoretical weight per linear meter, and n represents the total number of steel reinforcement detail records.

6. The bridge engineering quantity management method based on multi-terminal screenshots and WeChat push according to claim 1, characterized in that: In step (5), the WeChat scheduled push includes: setting the daily push time point, target group identifier, and message template ID; scheduling the scheduled task to trigger data retrieval according to the configured time; combining the retrieved material delivery data with the template engine to generate a text message, and calling the Excel generation library to create a material delivery form attachment; sending the message through the WeChat Webhook API, and recording the push status and results in the WeChat push log table.

7. The bridge engineering quantity management method based on multi-terminal screenshots and WeChat push according to claim 1, characterized in that: In step (6), when the user selects the material quantity deviation report template, the engineering quantity deviation rate is automatically calculated. : ； In the formula, This represents the actual amount of work done. For the design engineering quantity; when When this happens, it will be highlighted in the report using a preset marker.

8. An electronic device, comprising: processor; as well as A memory storing computer program instructions, which, when executed by the processor, cause the processor to perform the bridge engineering quantity management method based on multi-terminal screenshots and WeChat push as described in any one of claims 1-7.

9. A computer-readable storage medium having stored thereon computer program instructions, which, when executed by a processor, cause the processor to perform the bridge engineering quantity management method based on multi-terminal screenshots and WeChat push as described in any one of claims 1-7.

Images