CN-121998671-A - Building construction cost calculation quantity intelligent system and method based on large language model

CN121998671ACN 121998671 ACN121998671 ACN 121998671ACN-121998671-A

Abstract

The invention relates to the technical field of construction cost calculation, in particular to an intelligent system and method for building construction cost calculation based on a large language model. The system comprises a user-defined submenu module, a data acquisition and analysis module, a form file analysis module, a large language model matching module, an optimization and auditing module, an Internet collaboration and sharing module and a one-key quotation generating module. By constructing an end-to-end automatic system integrated with a special large language model intelligent matching core, automatic acquisition and standardization of multi-source heterogeneous data, intelligent project matching based on semantic understanding, compliance auditing and optimization of man-machine cooperation and complete process communication of final quotation file generation are realized, so that on the premise of keeping industry specification compliance, the processing efficiency and result accuracy of cost calculation are remarkably improved, and the adaptability of the system to globalization complex projects is enhanced.

Inventors

YUAN QIHUI
LV XIANGYU
LI ZHENGQUAN
WANG YUANTAO
YANG ZHIPENG
WANG XIAOLEI
HU YANJUN

Assignees

中国二十二冶集团有限公司

Dates

Publication Date: 20260508
Application Date: 20260127

Claims (10)

1. The building construction cost calculation intelligent system based on the large language model is characterized by comprising a custom submenu module, a data acquisition and analysis module, a table file analysis module, a large language model matching module, an optimization and auditing module, an Internet collaboration and sharing module and a one-key quotation generating module; The custom submenu module is respectively connected with the data acquisition and analysis module, the table file analysis module and the large language model matching module and is used for transmitting configuration parameters to the modules; The data acquisition and analysis module and the form file analysis module are respectively connected with the large language model matching module and are used for providing standardized project cost data and bid-inviting list data for the large language model matching module; The large language model matching module is connected with the optimizing and auditing module and is used for outputting a preliminary matching result to the optimizing and auditing module; the optimizing and auditing module is connected with the one-key quotation generating module and is used for transmitting quotation data confirmed by auditing to the one-key quotation generating module; The Internet collaboration and sharing module is respectively connected with the data acquisition and analysis module, the form file analysis module, the optimization and auditing module and the one-key quotation generating module and is used for providing a collaboration operation environment and data storage management service for the modules; The user-defined submenu module is provided with configuration submenu items for providing a visual configuration interface and intelligent input submenu items for providing a flow triggering function; the data acquisition and analysis module comprises a multi-source data acquisition unit for acquiring project cost management guide price files from multi-source data according to user configuration and a multi-language analysis and structuring processing unit for performing multi-language analysis and structuring processing on the acquired files; The form file analysis module comprises a file format adaptation unit and a key information extraction unit for adapting various formats of bid amounts and extracting key information, and a currency conversion unit for performing automatic currency conversion; The large language model matching module comprises a large language model building unit which is finely adjusted based on the construction industry cost calculation corpus and an intelligent matching unit which is used for executing intelligent matching and outputting a result; The optimizing and auditing module comprises an automatic optimizing unit for embedding an industry specification and an enterprise standard rule base and a manual auditing unit for providing a visual operation interface; The internet collaboration and sharing module comprises a collaborative work platform supporting multi-role authority management and real-time collaboration and a cloud storage and version management unit for realizing data cloud storage and version management; The one-key quotation generating module comprises a total price calculating unit for calculating the total quotation of the project and a list generating and deriving unit for generating and deriving a list of engineering quantities with prices.
2. The large language model based construction cost amount of calculation intelligent system of claim 1, wherein the configuration submenu item is provided with a parameter configuration interface for setting data acquisition sources, multilingual recognition preferences, currency conversion benchmarks, and data matching accuracy thresholds.
3. The large language model based construction cost accounting intelligent system of claim 1, wherein the multi-source data acquisition unit is provided with a first acquisition interface for downloading the guide price file from the enterprise integrated management system through an application programming interface calling mode, and a second acquisition interface for reading the guide price data from the local database through a structured query language.
4. The large language model based construction cost calculation intelligent system of claim 1, wherein the multi-language analysis and structuring processing unit integrates optical character recognition technology and multi-language natural language processing model, and is used for extracting project names, project features and unit of measure fields from unstructured files and converting the project names, the project features and the unit of measure fields into structuring data of a preset data model.
5. The large language model based construction cost calculation intelligent system according to claim 1, wherein the large language model construction unit is provided with a model training architecture for fine tuning a general large language model by adopting a low-rank adaptation technology, and the architecture is based on a construction industry cost calculation corpus comprising construction engineering quantity list pricing specifications, enterprise history matching cases and multilingual project feature description corpus.
6. The large language model based construction cost calculation intelligent system of claim 1, wherein the intelligent matching unit is configured with multi-level matching logic comprising matching weights of project names and measuring units and matching weights of project feature details, and the unit calculates total similarity of the bid listing project and the guide price project based on semantic understanding capability of the large language model.
7. The large language model based building construction cost calculation intelligent system of claim 1, wherein the cooperative work platform is provided with a right management component for distributing data operation rights based on user roles, a real-time cooperation component for realizing data synchronization by adopting a WebSocket technology, and a message notification component for pushing audit reminding and exception reminding.
8. A building construction cost calculation quantity intelligent method based on a large language model is characterized in that, use of the large language model-based building construction cost calculation intelligent system according to any one of claims 1 to 7, comprising the following specific steps: s1, receiving system parameters and a flow triggering instruction configured by a user through a custom submenu module; S2, automatically acquiring project cost management guidance price data through a data acquisition and analysis module, and performing multi-language analysis, structural conversion and verification on the data to generate standardized guidance price structural data; S3, reading a bid-posting list file through a form file analysis module, extracting key information, and executing automatic currency conversion when the fact that the price currency is inconsistent with the target currency is identified, so as to generate list structured data compatible with a guiding price data format; S4, calling a large language model subjected to corpus fine adjustment in the building field through a large language model matching module, performing intelligent matching on the list structured data and the guide price structured data according to a multi-level matching rule, and outputting a matching result and a similarity analysis report; S5, carrying out automatic specification compliance verification on the matching result through an optimization and auditing module, generating an optimization suggestion, and submitting manual auditing confirmation through a visual interface; s6, providing a real-time collaboration environment based on authority for the multi-role user through an Internet collaboration and sharing module, and encrypting and storing the data and operation log after verification and confirmation to a cloud server; And S7, automatically calculating the total quotation based on the finally confirmed unit price and the engineering quantity by a one-key quotation generating module, and generating a project quantity list file with prices which accords with the national standard.
9. The building construction cost calculation intelligent method based on the large language model of claim 8, wherein in step S2, multi-language analysis and structuring processing is carried out on the project cost management guide price file, specifically, the method comprises the steps of extracting texts and key fields by adopting an optical character recognition technology and a multi-language natural language processing model aiming at a PDF format file, automatically identifying header and data lines and extracting field information aiming at an Excel format file, and converting the extracted information into JSON format structured data according to a preset data model.
10. The method for intelligently calculating the construction cost based on the large language model is characterized in that in the step S4, intelligent matching is carried out according to a multi-level matching rule, wherein the method specifically comprises the steps of firstly matching a project name with a measuring unit to obtain first similarity, then matching project feature details to obtain second similarity, weighting and summing the first similarity and the second similarity according to preset weights to obtain total similarity, outputting project unit price successfully matched if the total similarity is not lower than a precision threshold configured by a user, and otherwise outputting a plurality of recommended items and difference analysis reports with the similarity ranked at the front.

Description

Building construction cost calculation quantity intelligent system and method based on large language model Technical Field The invention relates to the technical field of construction cost calculation, in particular to an intelligent system and method for building construction cost calculation based on a large language model. Background In the field of constructional engineering, the cost calculation is a core link of project investment control, bidding decision and construction cost management, and the accuracy directly influences the economic benefit and market competitiveness of the project. The traditional cost calculation work is highly dependent on manual operation, professionals need to manually compare a bid-drawing engineering quantity list with project cost guidance price files in enterprises, detailed information such as project names, technical characteristics, measuring units and the like is matched one by one, the process is tedious and time-consuming and tedious, particularly for large projects containing thousands of minutes, the manual processing period is always Zhou Ji, and the quick-response bid-drawing requirement is difficult to adapt. With the acceleration of globalization process in building industry, international engineering projects are increasing, and the cost and calculation work faces new challenges. On the one hand, the bidding documents possibly relate to multinational languages, the expression difference of the professional terms is easy to cause manual identification and translation deviation, and on the other hand, the real-time exchange rate conversion is needed for pricing of different currencies, so that the operation complexity and the error risk are further increased. Although partial cost auxiliary software exists in the market and can realize basic data importing and calculating functions, the method still relies on manual preprocessing when unstructured or multi-format files (such as PDF (portable document format) and complex Excel) are processed, and the method is mostly based on keyword face matching at an information matching layer, is difficult to cope with actual situations with the same semantics but different expressions (such as C30 commodity concrete and pre-mixed C30 concrete), and further lacks automatic analysis and intelligent conversion capability for multi-language texts. In recent years, artificial intelligence technology, particularly a large language model, has made remarkable progress in natural language understanding and cross-modal information processing, and a new technical path is provided for automation and intelligent application in the vertical field. However, how to deeply integrate such a front-edge technology with a specific business scene with strong specialization, high standardization and various data sources of construction engineering cost, and construct an intelligent calculation system capable of processing multi-source data end to end, realizing accurate semantic matching and supporting collaborative operation, is still a technical problem to be solved in the current industry. Disclosure of Invention Aiming at the defects of the prior art, the invention aims to provide a building construction cost calculation intelligent system and method based on a large language model, which can improve the cost calculation efficiency and quotation accuracy. The technical scheme adopted for solving the technical problems is as follows: The building construction cost calculation amount intelligent system based on the large language model comprises a custom submenu module, a data acquisition and analysis module, a form file analysis module, a large language model matching module, an optimization and auditing module, an Internet collaboration and sharing module and a one-key quotation generating module; The custom submenu module is respectively connected with the data acquisition and analysis module, the table file analysis module and the large language model matching module and is used for transmitting configuration parameters to the modules; The data acquisition and analysis module and the form file analysis module are respectively connected with the large language model matching module and are used for providing standardized project cost data and bid-inviting list data for the large language model matching module; The large language model matching module is connected with the optimizing and auditing module and is used for outputting a preliminary matching result to the optimizing and auditing module; the optimizing and auditing module is connected with the one-key quotation generating module and is used for transmitting quotation data confirmed by auditing to the one-key quotation generating module; The Internet collaboration and sharing module is respectively connected with the data acquisition and analysis module, the form file analysis module, the optimization and auditing module and the one-key quotation generating module and is used for