文章详情

题名 · 作者 · 摘要 · 关键词 · DOI · 参考文献

面向ROS知识问答的层次化图检索增强生成方法

施亚龙,赵俊,侯劲

(四川轻化工大学,四川 宜宾 644005)

栏目:计算机技术 · 第 2026 年第 17 期 · 页码:58-65 · 2096-4706(2026)17-0058-08 · 中图分类号:TP181;TP391.1 · 文献标识码:A · DOI:10.19850/j.cnki.2096-4706.2026.17.011

摘要

针对机器人操作系统(ROS)图检索过程中易出现的上下文碎片化、宏微观结构脱节问题,本文提出一种层次化图检索增强生成框架(HiGraphRAG)。该方法通过全局图谱重组修复物理连通断层,依托高斯混合模型构建自底向上的层次化摘要索引,并结合视角扩展机制实现多粒度结构化知识的精准召回。基于自制ROS数据集的实验结果表明,相较于主流GraphRAG范式,HiGraphRAG的总体生成质量胜率达56.4%,知识多样性胜率达60.2%。该框架可有效弥补系统信息断层,显著提升大模型复杂推理的全面性与知识赋能能力。

关键词

检索增强生成;知识图谱;机器人操作系统;层次化索引;层次化检索

参考文献

  1. [1] Lewis P,Perez E,Piktus A,et al. Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks [C]//Advances in Neural Information Processing Systems(S1049-5258),2020,33:9459-9474.
  2. [2] Gao Y F,Xiong Y,Gao X Y,et al. Retrieval-Augmented Generation for Large Language Models: A Survey [PP/OL].arXiv:2312.10997 [cs.CL].https://doi.org/10.48550/arXiv.2312.10997.
  3. [3] Christmann P,Weikum G. RAG-Based Question Answering over Heterogeneous Data and Text [PP/OL].arXiv:2412.07420 [cs.CL].https://doi.org/10.48550/arXiv.2412.07420.
  4. [4] Nguyen A T P,Hoang V D. Development of Code Evaluation System Based on Abstract Syntax Tree [J].Journal of Technical Education Science(S1859-1272),2024,19(Special Issue 01):15-24.
  5. [5] Wang Y C,Guo S X,Tan C W.From Code Generation to Software Testing: AI Copilot With Context-Based Retrieval-Augmented Generation [J].IEEE Software,2025,42(4):34-42.
  6. [6] 奥帅,王嘉欢,马波.基于知识图谱关系路径推理及相似度判别的机械故障源精确定位方法 [J].机电工程,2025,42(2):387-396.
  7. [7] Li M Z,Yang C H,Xu C J,et al. Context-Aware Inductive Knowledge Graph Completion with Latent Type Constraints and Subgraph Reasoning [C]//Proceedings of the AAAI Conference on Artificial Intelligence(S2159-5399),2025,39(11):12102-12111.
  8. [8] Zhu X R,Xie Y X,Liu Y, et al. Knowledge Graph-Guided Retrieval Augmented Generation [C]//Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies,2025:8912-8924.
  9. [9] Kim S,Hwang S J,Kim J H,et al. ReGraphRAG: Reorganizing Fragmented Knowledge Graphs for Multi-Perspective Retrieval-Augmented Generation [C]//Findings of the Association for Computational Linguistics: EMNLP 2025,2025:5426-5443.
  10. [10] Edge D,Trinh H,Cheng N,et al. From Local to Global: A Graph RAG Approach to Query-Focused Summarization [PP/OLJ].arXiv:2404.16130 [cs.CL].https://doi.org/10.48550/arXiv.2404.16130.
  11. [11] Huang H,Huang Y,Yang J,et al. Retrieval-Augmented Generation with Hierarchical Knowledge [PP/OL].arXiv:2503.10150 [cs.CL].https://doi.org/10.48550/arXiv.2503.10150.
  12. [12] 赵衍,程恺.基于检索增强生成技术的规章制度问答大语言模型的构 [J].计算机应用与软件,2025,42(7):192-198.
  13. [13] 翁嘉屹,翟洁,熊瀚锐,等.基于知识图谱和RAG技术的古诗词智能问答系 [J].现代信息科技,2025,9(20):115-120,125.
  14. [14] 王洪彪,赵继春,陈蕾.基于知识图谱的动物疾病智能问答系统设计与实 [J].计算机应用与软件,2025,42(12):29-34,40.
  15. [15] 贺梓然,江波,王晓龙.改进检索增强与LLM思维链维修策略生 [J].计算机应用与软件,2025,42(3):1-6,83.
  16. [16] 许智宏,谭金鸽,王利琴,等.融合关系层次结构的知识图谱嵌 [J].计算机应用与软件,2024,41(2):152-157,228.
  17. [17] Guo Z R,Xia L H,Yu Y H,et al. LightRAG: Simple and Fast Retrieval-Augmented Generation [PP/OL].arXiv:2410.05779 [cs.IR].https://doi.org/10.48550/arXiv.2410.05779.
  18. [18] Frana P L,Misa T J. An Interview with Edsger W. Dijkstra [J].Communications of the ACM(S0001-0782),2010,53(8):41-47.
  19. [19] Ehrmann M,Hamdi A,Pontes E L,et al. Named Entity Recognition and Classification in Historical Documents: A Survey [J].ACM Computing Surveys(S0360-0300),2023,56(2):1-28.

作者简介

施亚龙(1999—),男,汉族,湖北宜昌人,硕士研究生在读,研究方向:大模型;通信作者:赵俊(1984—),男,汉族,四川自贡人,副教授,博士研究生,研究方向:网络化控制系统、多智能体时变编队控制;侯劲(1968—),男,汉族,四川自贡人,教授,博士研究生,研究方向:人工智能、智能检测。