Empowering knowledge graphs with hybrid retrieval-augmented generation for the intelligent mix scheme of mass concrete

To achieve precise multi-objective mix design of dam concrete-targeting strength, durability, and crack resistance-this study proposes a Dam Concrete Mix Proportioning intelligent generation System (DCMPS) based on a Retrieval-Augmented Generation (RAG) framework. Firstly, a total of 1723 mix propor...

Full description

Saved in:
Bibliographic Details
Main Authors: Yuqing Shang, Zhijiang Ke, Peng Lin, Qianhui Ren, Wenshan Zhang, Xiongwu Wang, Xiaotao Li, Fuyuan Gong, Shiqi Wang, Baofa Wang, Zhengkui Xu, Minglun Sun, Shunli Tan
Format: Article
Language:English
Published: Elsevier 2025-12-01
Series:Case Studies in Construction Materials
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2214509525007776
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850120646796771328
author Yuqing Shang
Zhijiang Ke
Peng Lin
Qianhui Ren
Wenshan Zhang
Xiongwu Wang
Xiaotao Li
Fuyuan Gong
Shiqi Wang
Baofa Wang
Zhengkui Xu
Minglun Sun
Shunli Tan
author_facet Yuqing Shang
Zhijiang Ke
Peng Lin
Qianhui Ren
Wenshan Zhang
Xiongwu Wang
Xiaotao Li
Fuyuan Gong
Shiqi Wang
Baofa Wang
Zhengkui Xu
Minglun Sun
Shunli Tan
author_sort Yuqing Shang
collection DOAJ
description To achieve precise multi-objective mix design of dam concrete-targeting strength, durability, and crack resistance-this study proposes a Dam Concrete Mix Proportioning intelligent generation System (DCMPS) based on a Retrieval-Augmented Generation (RAG) framework. Firstly, a total of 1723 mix proportioning records collected from 15 hydropower stations were structurally extracted using a customized data transformation tool. A knowledge graph (KG) was then constructed on the Neo4j platform, encompassing entities such as environmental conditions, material compositions, and performance indicators. Secondly, the Moka Massive Mixed Embedding (M3E) model was employed to encode the KG into a vector database, enabling accurate retrieval of relevant mix proportioning cases via cosine similarity calculation. Finally, based on a hierarchical reasoning architecture driven by prompt engineering, DCMPS retrieves relevant cases from the vector database using a semantic search mechanism. A large language model (LLM) conducts comparative analysis on the retrieved cases to generate mix design schemes tailored to specific requirements. The DCMPS’s performance was validated using a test dataset, The results demonstrate that the system outperforms the larger-parameter native LLM in overall evaluation on the test dataset, confirming that knowledge augmentation significantly enhances the performance of smaller models. Case studies indicate that the recommended mix designs closely align with the target performance requirements, and the comparison of relevant cases along with the analysis of material influence mechanisms enhances the interpretability of the results. This DCMPS provides effective technical support for intelligent analysis and engineering decision-making in concrete mix proportioning.
format Article
id doaj-art-75f1c968fef44cfaba6b8525f5e4ecb0
institution OA Journals
issn 2214-5095
language English
publishDate 2025-12-01
publisher Elsevier
record_format Article
series Case Studies in Construction Materials
spelling doaj-art-75f1c968fef44cfaba6b8525f5e4ecb02025-08-20T02:35:19ZengElsevierCase Studies in Construction Materials2214-50952025-12-0123e0497910.1016/j.cscm.2025.e04979Empowering knowledge graphs with hybrid retrieval-augmented generation for the intelligent mix scheme of mass concreteYuqing Shang0Zhijiang Ke1Peng Lin2Qianhui Ren3Wenshan Zhang4Xiongwu Wang5Xiaotao Li6Fuyuan Gong7Shiqi Wang8Baofa Wang9Zhengkui Xu10Minglun Sun11Shunli Tan12School of Civil Engineering and Water Resources, Qinghai University, Xining 810016, China; Sichuan Energy Internet Research Institute, Tsinghua University, Chengdu 610213, ChinaSchool of Civil Engineering and Water Resources, Qinghai University, Xining 810016, China; Sichuan Energy Internet Research Institute, Tsinghua University, Chengdu 610213, ChinaSchool of Civil Engineering and Water Resources, Qinghai University, Xining 810016, China; Department of Hydraulic Engineering, Tsinghua University, Beijing 100084, China; Sichuan Energy Internet Research Institute, Tsinghua University, Chengdu 610213, China; Corresponding author at: School of Civil Engineering and Water Resources, Qinghai University, Xining 810016, China.School of Civil Engineering and Water Resources, Qinghai University, Xining 810016, ChinaSinohydro Engineering Bureau No.4 Co., Ltd., Xining 810007, ChinaSinohydro Engineering Bureau No.4 Co., Ltd., Xining 810007, ChinaSinohydro Engineering Bureau No.4 Co., Ltd., Xining 810007, ChinaCollege of Architectural Engineering, Zhejiang University, Hangzhou 310058, ChinaCollege of Architectural Engineering, Zhejiang University, Hangzhou 310058, ChinaSinohydro Engineering Bureau No.4 Co., Ltd., Xining 810007, ChinaSinohydro Engineering Bureau No.4 Co., Ltd., Xining 810007, ChinaChina Three Gorges Construction Engineering Corporation, Chengdu 610042, ChinaChina Water Conservancy and Hydropower Engineering Bureau No.11 Co., Ltd., Zhengzhou 450001, ChinaTo achieve precise multi-objective mix design of dam concrete-targeting strength, durability, and crack resistance-this study proposes a Dam Concrete Mix Proportioning intelligent generation System (DCMPS) based on a Retrieval-Augmented Generation (RAG) framework. Firstly, a total of 1723 mix proportioning records collected from 15 hydropower stations were structurally extracted using a customized data transformation tool. A knowledge graph (KG) was then constructed on the Neo4j platform, encompassing entities such as environmental conditions, material compositions, and performance indicators. Secondly, the Moka Massive Mixed Embedding (M3E) model was employed to encode the KG into a vector database, enabling accurate retrieval of relevant mix proportioning cases via cosine similarity calculation. Finally, based on a hierarchical reasoning architecture driven by prompt engineering, DCMPS retrieves relevant cases from the vector database using a semantic search mechanism. A large language model (LLM) conducts comparative analysis on the retrieved cases to generate mix design schemes tailored to specific requirements. The DCMPS’s performance was validated using a test dataset, The results demonstrate that the system outperforms the larger-parameter native LLM in overall evaluation on the test dataset, confirming that knowledge augmentation significantly enhances the performance of smaller models. Case studies indicate that the recommended mix designs closely align with the target performance requirements, and the comparison of relevant cases along with the analysis of material influence mechanisms enhances the interpretability of the results. This DCMPS provides effective technical support for intelligent analysis and engineering decision-making in concrete mix proportioning.http://www.sciencedirect.com/science/article/pii/S2214509525007776Mixture designKnowledge graphLarge language modelRetrieval-augmented generationSemantic retrieval
spellingShingle Yuqing Shang
Zhijiang Ke
Peng Lin
Qianhui Ren
Wenshan Zhang
Xiongwu Wang
Xiaotao Li
Fuyuan Gong
Shiqi Wang
Baofa Wang
Zhengkui Xu
Minglun Sun
Shunli Tan
Empowering knowledge graphs with hybrid retrieval-augmented generation for the intelligent mix scheme of mass concrete
Case Studies in Construction Materials
Mixture design
Knowledge graph
Large language model
Retrieval-augmented generation
Semantic retrieval
title Empowering knowledge graphs with hybrid retrieval-augmented generation for the intelligent mix scheme of mass concrete
title_full Empowering knowledge graphs with hybrid retrieval-augmented generation for the intelligent mix scheme of mass concrete
title_fullStr Empowering knowledge graphs with hybrid retrieval-augmented generation for the intelligent mix scheme of mass concrete
title_full_unstemmed Empowering knowledge graphs with hybrid retrieval-augmented generation for the intelligent mix scheme of mass concrete
title_short Empowering knowledge graphs with hybrid retrieval-augmented generation for the intelligent mix scheme of mass concrete
title_sort empowering knowledge graphs with hybrid retrieval augmented generation for the intelligent mix scheme of mass concrete
topic Mixture design
Knowledge graph
Large language model
Retrieval-augmented generation
Semantic retrieval
url http://www.sciencedirect.com/science/article/pii/S2214509525007776
work_keys_str_mv AT yuqingshang empoweringknowledgegraphswithhybridretrievalaugmentedgenerationfortheintelligentmixschemeofmassconcrete
AT zhijiangke empoweringknowledgegraphswithhybridretrievalaugmentedgenerationfortheintelligentmixschemeofmassconcrete
AT penglin empoweringknowledgegraphswithhybridretrievalaugmentedgenerationfortheintelligentmixschemeofmassconcrete
AT qianhuiren empoweringknowledgegraphswithhybridretrievalaugmentedgenerationfortheintelligentmixschemeofmassconcrete
AT wenshanzhang empoweringknowledgegraphswithhybridretrievalaugmentedgenerationfortheintelligentmixschemeofmassconcrete
AT xiongwuwang empoweringknowledgegraphswithhybridretrievalaugmentedgenerationfortheintelligentmixschemeofmassconcrete
AT xiaotaoli empoweringknowledgegraphswithhybridretrievalaugmentedgenerationfortheintelligentmixschemeofmassconcrete
AT fuyuangong empoweringknowledgegraphswithhybridretrievalaugmentedgenerationfortheintelligentmixschemeofmassconcrete
AT shiqiwang empoweringknowledgegraphswithhybridretrievalaugmentedgenerationfortheintelligentmixschemeofmassconcrete
AT baofawang empoweringknowledgegraphswithhybridretrievalaugmentedgenerationfortheintelligentmixschemeofmassconcrete
AT zhengkuixu empoweringknowledgegraphswithhybridretrievalaugmentedgenerationfortheintelligentmixschemeofmassconcrete
AT minglunsun empoweringknowledgegraphswithhybridretrievalaugmentedgenerationfortheintelligentmixschemeofmassconcrete
AT shunlitan empoweringknowledgegraphswithhybridretrievalaugmentedgenerationfortheintelligentmixschemeofmassconcrete