CN-121981788-A - Pipe network engineering cost estimation method and system based on AI

CN121981788ACN 121981788 ACN121981788 ACN 121981788ACN-121981788-A

Abstract

The application relates to the technical field of pipe network engineering cost estimation based on AI, and discloses a pipe network engineering cost estimation method and a pipe network engineering cost estimation system based on AI, wherein the method comprises the steps of firstly analyzing basic engineering parameters of pipe network engineering to be estimated and extracting engineering constraint conditions based on a preset cost estimation rule file; the method comprises the steps of determining a rule applicable path according to constraint conditions and rule applicable sequences, constructing a corresponding prompt language chain which comprises a plurality of prompt language units split in sequence, sequentially executing engineering calculation subtasks corresponding to the prompt language units by using a large language model to generate a cost estimation intermediate result, carrying out consistency check including monotonicity check on the intermediate result, and outputting the estimation result when the consistency check meets preset conditions. The application combines the AI technology and the regularization flow, improves the efficiency and the accuracy of pipe network engineering cost estimation, solves the problems of low efficiency and insufficient accuracy of the traditional method, and is suitable for the early-stage cost estimation of various pipe network engineering.

Inventors

Dai Meifei
XIONG LEI
HUANG DEYUAN
HE YANPING
ZHANG WENJIAN
ZHANG XIN
LIU MUJUN

Assignees

广州市市政工程设计研究总院有限公司

Dates

Publication Date: 20260505
Application Date: 20260123

Claims (10)

1. The pipe network engineering cost estimation method based on the AI is characterized by comprising the following steps: analyzing basic engineering parameters of pipe network engineering to be estimated, and extracting engineering constraint conditions from the basic engineering parameters based on a preset cost estimation rule file; Determining a rule applicable path matched with the current pipe network engineering condition to be estimated based on the engineering constraint condition and a predefined rule applicable sequence in the preset cost estimation rule file; constructing a prompt language chain corresponding to the rule applicable path according to the determined rule applicable path, wherein the prompt language chain comprises a plurality of prompt language units which are split according to the rule applicable sequence; sequentially executing engineering calculation subtasks corresponding to the prompt units according to the arrangement sequence of the prompt units in the prompt chain by using a large language model, and generating a pipe network engineering cost estimation intermediate result; carrying out consistency check on the construction cost estimation intermediate result, wherein the consistency check at least comprises checking the monotonicity of the estimation result along with the change of engineering parameters; And outputting a pipe network engineering cost estimation result when the consistency check meets the preset condition.
2. The AI-based pipe network engineering cost estimation method of claim 1, wherein, The basic engineering parameters at least comprise pipe diameter and pipe burial depth, and the engineering constraint conditions at least comprise groove support mode applicable conditions determined according to the corresponding relation between the pipe burial depth and the support mode in the construction cost estimation rule file and index interval dividing conditions determined according to pipe diameter interval dividing standards in the pipe diameter and construction cost estimation rule file.
3. The AI-based pipe network engineering cost estimation method of claim 2, wherein, And when the pipe diameter or the pipe burial depth of the pipeline in the basic engineering parameters is positioned at the boundary of the corresponding divided interval, marking the pipeline diameter or the pipe burial depth as an interval boundary working condition, and extracting constraint conditions needing to introduce index parameters of adjacent intervals as alternative paths.
4. The AI-based pipe network engineering cost estimation method of claim 3, wherein, The rule applicable path is determined by firstly determining the category of an applicable cost estimation index table, secondly determining a main applicable section according to the index section dividing condition and the groove supporting mode applicable condition, and introducing a section adjacent to the main applicable section as an alternative path when the section boundary working condition is marked.
5. The AI-based pipe network engineering cost estimation method of claim 4, wherein, The construction of the prompt language chain comprises splitting a pipe network engineering cost estimation task into a plurality of engineering calculation subtasks, wherein each subtask corresponds to one prompt language unit, the prompt language unit comprises a unit for reading and confirming engineering parameters, a unit for locating a target position in a cost estimation index table, a unit for reading index parameters and a unit for carrying out cost calculation, and when the rule applicable path comprises an alternative path, the prompt language chain further comprises a prompt language unit for processing the index parameters of the alternative path and calculating.
6. The AI-based pipe network engineering cost estimation method of claim 5, wherein, The engineering calculation sub-task at least comprises the steps of positioning a target line number and a target column number in the cost estimation index table according to the rule applicable path, reading a digging and filling support coefficient and a buried pipe index corresponding to the target line number and the target column number, and calculating the basic cost of the pipeline according to a preset formula based on the digging and filling support coefficient and the buried pipe index.
7. The AI-based pipe network engineering cost estimation method of claim 6, wherein, When the rule applicable path comprises an alternative path, executing the engineering calculation subtask further comprises the steps of respectively calculating the pipeline basic cost corresponding to the main applicable section and the alternative path, and performing linear interpolation calculation according to interpolation logic agreed by the cost estimation rule file to obtain the final pipeline basic cost.
8. The AI-based pipe network engineering cost estimation method of claim 1, wherein, The consistency check further comprises checking whether the estimated result has abnormal mutation among different partition intervals and checking whether the result calculated by adopting a linear interpolation method accords with interpolation logic.
9. An AI-based pipe network engineering cost estimation system, comprising: The parameter analysis and condition extraction module analyzes basic engineering parameters of the pipe network engineering to be estimated, and extracts engineering constraint conditions from the basic engineering parameters based on a preset cost estimation rule file; The rule path determining module is used for determining a rule applicable path matched with the current pipe network engineering condition to be estimated based on the engineering constraint condition and a predefined rule applicable sequence in the preset cost estimation rule file; the prompt language chain construction module constructs a prompt language chain corresponding to the rule application path according to the determined rule application path, wherein the prompt language chain comprises a plurality of prompt language units which are split according to the rule application sequence; The large language model processing module is used for sequentially executing engineering calculation subtasks corresponding to the prompt units according to the arrangement sequence of the prompt units in the prompt chain by utilizing a large language model, and generating a pipe network engineering cost estimation intermediate result; the consistency check module is used for carrying out consistency check on the construction cost estimation intermediate result, and the consistency check at least comprises check on the monotonicity of the estimation result along with the change of engineering parameters; And the result output module is used for outputting a pipe network engineering cost estimation result when the consistency check meets the preset condition.
10. A computer device, the computer device comprising: and a memory communicatively coupled to the at least one processor, wherein, The memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-8.

Description

Pipe network engineering cost estimation method and system based on AI Technical Field The application relates to the technical field of pipe network engineering cost estimation based on AI, in particular to a pipe network engineering cost estimation method and system based on AI. Background The pipe network engineering is used as a core component of municipal infrastructure construction and covers a plurality of fields of water supply and drainage, fuel gas, heating power and the like, and the cost estimation result directly influences the scientificity of project standing decision, fund configuration and scheme optimization. The current pipe network engineering cost estimation is mainly finished by manually consulting rated standard and applying index parameters, and the method has obvious technical bottlenecks, on one hand, the estimation process is required to manually match complex engineering constraint conditions such as the corresponding relation between pipe diameter and buried depth and supporting mode and index interval, so that the efficiency is low, estimation deviation is easily caused by manual misoperation, and on the other hand, the traditional method only adopts single interval parameters for calculation aiming at the special working condition that the pipe diameter and buried depth are at interval boundaries, and cannot be attached to the transition characteristics of engineering actual cost, so that the boundary working condition estimation precision is insufficient. Meanwhile, the existing AI auxiliary estimation technology has the problems of irregular model input and fuzzy task splitting, is difficult to accurately execute the cost calculation subtasks, lacks a systematic result checking mechanism, and cannot effectively identify logic contradictions such as monotonicity abnormality, interval mutation and the like in the estimation result. In addition, the traditional method has no clear correction path when the estimation error occurs, and only the whole process can be re-executed, so that the estimation efficiency and reliability are further reduced. These problems lead to the difficulty in meeting the requirements of rapid, accurate and standard estimation in the early stage of engineering design in the existing pipe network engineering cost estimation method. Disclosure of Invention The application aims to provide an AI-based pipe network engineering cost estimation method and system, which are used for solving the problems in the background technology. According to one aspect of the present application, there is provided an AI-based pipe network engineering cost estimation method, including the steps of: analyzing basic engineering parameters of pipe network engineering to be estimated, and extracting engineering constraint conditions from the basic engineering parameters based on a preset cost estimation rule file; Determining a rule applicable path matched with the current pipe network engineering condition to be estimated based on the engineering constraint condition and a predefined rule applicable sequence in the preset cost estimation rule file; constructing a prompt language chain corresponding to the rule applicable path according to the determined rule applicable path, wherein the prompt language chain comprises a plurality of prompt language units which are split according to the rule applicable sequence; sequentially executing engineering calculation subtasks corresponding to the prompt units according to the arrangement sequence of the prompt units in the prompt chain by using a large language model, and generating a pipe network engineering cost estimation intermediate result; carrying out consistency check on the construction cost estimation intermediate result, wherein the consistency check at least comprises checking the monotonicity of the estimation result along with the change of engineering parameters; And outputting a pipe network engineering cost estimation result when the consistency check meets the preset condition. The basic engineering parameters at least comprise pipe diameter and pipe burial depth, and the engineering constraint conditions at least comprise groove support mode applicable conditions determined according to the corresponding relation between the pipe burial depth and the support modes in the construction cost estimation rule file and index interval dividing conditions determined according to pipe diameter interval dividing standards in the pipe diameter and the construction cost estimation rule file. Preferably, when the pipe diameter or the pipe burial depth in the basic engineering parameters is at the boundary of the corresponding division interval, the boundary working condition is marked, and the constraint condition that the index parameters of the adjacent interval need to be introduced is extracted as the constraint condition of the alternative path. Preferably, determining the rule applicable path comprises firstly determining a