铁道标准设计2025，Vol.69Issue(3)：73-82,10.DOI:10.13238/j.issn.1004-2954.202407010003
知识图谱驱动的关键结构物工程方案智能决策关键技术研究与应用
Research and Application of Key Technologies for Knowledge Graph-Driven Intelligent Decision-Making in Critical Structural Engineering Solutions
摘要
Abstract
This study addresses the issues in the design of critical structures such as subgrades,bridges,and tunnels in land transportation engineering.Decision-making in these designs relies on subjective experience and lacks automation.A solution based on knowledge graphs and intelligent decision-making technologies was proposed.Knowledge graph construction techniques were employed to efficiently manage unstructured regulatory text data,and a method for knowledge graph construction applicable to subgrade,bridge,and tunnel design was proposed.Based on this,intelligent decision-making technologies for subgrade,bridge,and tunnel design schemes,driven by knowledge graphs,were researched and developed.Finally,through survey analysis,technical experiments,software development,and case verification,an intelligent decision-making system for subgrade,bridge,and tunnel design was developed and applied.The study results indicated that:(1)in subgrade design,by utilizing data dimensionality reduction,keyword association analysis,and machine learning methods,a matrix for screening design conditions and schemes was established.Joint decision-making for design schemes was achieved using graph neural networks and deep learning.(2)In bridge design,an integrated ontology-enhanced graph neural network,"MLP+GRU+GCN+Attention"(AGOAM model),was proposed.This included the development of an intelligent bridge span layout algorithm and a proactive notification system for bridge regulation clauses,enabling intelligent decision-making for bridge span layout and structural schemes.(3)In tunnel design,knowledge graph technology,and deep learning methods were applied to achieve intelligent decision-making for tunnel portal design parameters,while BIM technology was utilized for 3D visualization.By constructing knowledge graphs in the field of subgrade,bridge,and tunnel design and combining them with graph neural networks and deep learning technologies,this study developed a knowledge graph-based intelligent decision-making system for design schemes.This system significantly improves the quality and efficiency of design decision-making,thus providing robust technical support for the intelligent development of land transportation engineering design.关键词
陆路交通/知识图谱/关键结构物/工程方案/智能决策Key words
land transportation/knowledge graph/critical structures/engineering solutions/intelligent decision-making分类
交通运输引用本文复制引用
宋文祥,姚洪锡,钟晶,向子南,谢浩,柏华军,吴佳明..知识图谱驱动的关键结构物工程方案智能决策关键技术研究与应用[J].铁道标准设计,2025,69(3):73-82,10.基金项目
国家重点研发计划项目(2021YFB2600400) （2021YFB2600400）
中国铁建股份有限公司科技研发计划项目(2022-A02) （2022-A02）
