A knowledge graph-based engineering change cost rapid evaluation method

By constructing an engineering knowledge graph based on a knowledge graph approach and utilizing bidirectional graph neural networks and dynamic hierarchical graph attention networks, the problem of insufficient utilization of multi-source heterogeneous data in existing technologies is solved, enabling accurate estimation and transparent assessment of engineering change costs, and improving the accuracy and adaptability of the assessment.

CN122335382APending Publication Date: 2026-07-03SHANXI RUNSHIHE PROJECT MANAGEMENT CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202610430786.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-04-02
Publication Date
2026-07-03

AI Technical Summary

Technical Problem

Existing engineering change assessment methods are unable to fully utilize multi-source heterogeneous data, resulting in low cost estimation accuracy, poor adaptability, and a lack of transparency and credibility in the assessment results.

Method used

A knowledge graph-based approach is adopted, which constructs an engineering knowledge graph, uses a bidirectional graph neural network retrieval algorithm to identify the subgraphs affecting changes, and combines a dynamic hierarchical graph attention network with a spatiotemporal decay dependency mechanism to perform cost mapping and nonlinear correction, thereby achieving in-depth analysis and refined estimation of the impact of changes.

Benefits of technology

It significantly improves the accuracy and adaptability of engineering change cost assessment, enables comprehensive identification of the scope of change impact and refined cost calculation, and enhances the credibility and transparency of assessment results.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122335382A_ABST
    Figure CN122335382A_ABST
Patent Text Reader

Abstract

This invention discloses a rapid cost assessment method for engineering changes based on a knowledge graph, comprising: S1, synchronously collecting multi-source heterogeneous data for entity extraction and relation alignment to construct an engineering knowledge graph; S2, inputting change instructions, using a bidirectional graph neural network to identify initial nodes and generate a change impact subgraph; S3, introducing an improved dynamic hierarchical graph attention network with a spatiotemporal decay dependency mechanism, constructing a dynamic decay factor to weight and encode features at each level, and outputting a change feature vector; S4, calling quota standards for cost mapping, performing difference calculation and automated splitting, and outputting a basic cost dataset; S5, combining historical change database similarity matching, and outputting corrected estimated values ​​after deviation calibration and nonlinear correction; S6, encapsulating and generating an assessment report and writing it back to the knowledge graph archive. This invention achieves deep analysis of change propagation paths and refined dynamic cost estimation, improving assessment efficiency and accuracy.
Need to check novelty before this filing date? Find Prior Art