大语言模型在建筑工程中的应用测试与讨论

覃思中; 郑哲; 顾燚; 陆新征

doi:10.13204/j.gyjzG23081006

大语言模型在建筑工程中的应用测试与讨论

doi: 10.13204/j.gyjzG23081006

清华大学土木工程系, 北京 100084

基金项目:

清华-清尚智慧场景创新设计研究院2023年公开遴选项目；中华人民共和国住房和城乡建设部科学技术计划项目（2022-K-073）。

详细信息

作者简介:
覃思中,男,2001年出生,博士研究生。

通讯作者:
陆新征,luxz@tsinghua.edu.cn。

计量
- 文章访问数: 517
- HTML全文浏览量: 185
- PDF下载量: 34
- 被引次数: 0
出版历程
- 收稿日期: 2023-08-10
- 网络出版日期: 2023-11-08

Exploring and Discussion on the Application of Large Language Models in Construction Engineering

Department of Civil Engineering, Tsinghua University, Beijing 100084, China

摘要

摘要: 建筑业作为我国的经济支柱行业之一，一直面临着生产效率低、智能化水平有限等问题，大语言模型则给行业的发展带来了新的可能。首先提出了一套大语言模型在建筑工程中的应用方案，采用提示词工程和本地知识库相结合的方式来提升模型性能，通过实验分析验证其效果，并探究了本方案在行业各个领域中应用的可行性，针对部分任务提供了详细的应用案例。从实验结果中可以看出，尽管目前大语言模型在一些复杂的问题上还有较大的提升空间，但已经能初步替代建筑工程中的一些文本任务，为建筑业未来的发展提供了一个新的方向。
- 大语言模型 /
- 建筑工程 /
- 提示词工程 /
- 本地知识库 /
- 问答系统
Abstract: As one of China's key industries and economic pillars, the construction industry has long been plagued by low productivity and limited levels of automation. However, large language models present new possibilities for industry advancement. This paper proposes an application framework for large language models in construction engineering, utilizing prompt engineering and a local knowledge base to enhance model performance. The effectiveness of the proposed framework is validated through experimental analysis, exploring its feasibility in various domains within the industry and providing detailed application examples for specific tasks. The experimental results indicate that although there is still room for improvement in tackling complex problems, large language models have already demonstrated their potential to replace certain text-related tasks in construction engineering, offering a new direction for the future development of the construction industry.
- large language model /
- construction engineering /
- prompt engineering /
- local knowledge base /
- question-answering system

HTML全文

参考文献(36)

[1]	GOOGLE. 体验Bard-Google的AI实验项目[EB/OL].[2023-07-20]. https://bard.google.com.
[2]	DU Z, QIAN Y, LIU X, et al. GLM:General language model pretraining with autoregressive blank infilling[C]//Proceedings of the 60th annual meeting of the association for computational linguistics. 2022:320-335.
[3]	百度. 文心大模型-产业级知识增强大模型[EB/OL].[2023-07-20]. https://wenxin.baidu.com/.
[4]	阿里巴巴. 通义千问[EB/OL].[2023-08-09]. https://qianwen.aliyun.com/.
[5]	科大讯飞. 讯飞星火认知大模型[EB/OL].[2023-07-20]. https://xinghuo.xfyun.cn/.
[6]	赵峰, 王要武, 金玲, 等. 2022年建筑业发展统计分析[J]. 工程管理学报, 2023, 37(1):1-6.
[7]	许宪春, 王洋, 唐雅. 2022年中国经济形势分析与2023年展望[J]. 经济学动态, 2023(2):19-32.
[8]	加快建筑业转型推动高质量发展:住房和城乡建设部建筑市场监管司副司长廖玉平解读《指导意见》[J]. 工程建设标准化, 2020(8):12-14.
[9]	陆新征, 廖文杰, 顾栋炼, 等. 从基于模拟到基于人工智能的建筑结构设计方法研究进展[J/OL]. 工程力学:1-18[2023-09-17 ]. http://kns.cnki.net/kcms/detail/11.2595.O3.20230117.0853.002.html.
[10]	丁烈云, 徐捷, 覃亚伟. 建筑3D打印数字建造技术研究应用综述[J]. 土木工程与管理学报, 2015, 32(3):1-10.
[11]	郭红领, 王尧, 马琳瑶, 等. 土木工程施工安全研究的现状与趋势[J]. 华中科技大学学报(自然科学版), 2022, 50(8):89-98.
[12]	JR B F S, HOSKERE V, NARAZAKI Y. Advances in computer vision-based civil infrastructure inspection and monitoring[J]. Engineering, 2019, 5(2):199-248.
[13]	BAEK S, JUNG W, HAN S H. A critical review of text-based research in construction:data source, analysis method, and implications[J/OL]. Automation in Construction, 2021, 132,103915. https://doi.org/10.1016/j.autcon.2021.103915.
[14]	王煜, 邓晖, 李晓瑶, 等. 自然语言处理技术在建筑工程中的应用研究综述[J]. 图学学报, 2020, 41(4):501-511.
[15]	刘湧泉. 我国机器翻譯工作的进展[J]. 科学通报, 1959(17):563-564.
[16]	DING Y, MA J, LUO X. Applications of natural language processing in construction[J/OL]. Automation in Construction, 2022, 136. https://doi.org/10.1016/j.autcon.2022.104169.
[17]	CALDAS C H, SOIBELMAN L, HAN J. Automated classification of construction project documents[J]. Journal of Computing in Civil Engineering, 2002, 16(4):234-243.
[18]	CALDAS C H, SOIBELMAN L. Automating hierarchical document classification for construction management information systems[J]. Automation in Construction, 2003, 12(4):395-406.
[19]	KRIZHEVSKY A, SUTSKEVER I, HINTON G E. ImageNet classification with deep convolutional neural networks[C/OL]//Advances in Neural Information Processing Systems. Curran Associates, Inc., 2012[2023-08-04]. https://proceedings.neurips.cc/paper/2012/hash/c399862d3b9d6b76c8436e924a68c45b-Abstract.html.
[20]	ALOM M Z, TAHA T M, YAKOPCIC C, et al. The History Began from Alexnet:A Comprehensive Survey on Deep Learning Approaches[M/OL]. arXiv, 2018[2023-08-04]. http://arxiv.org/abs/1803.01164. https://doi.org/10.48550/arXiv.1803.01164.
[21]	YU W der, HSU J Y. Content-based text mining technique for retrieval of CAD documents[J]. Automation in Construction, 2013, 31:65-74.
[22]	SHEN L, YAN H, FAN H, et al. An integrated system of text mining technique and case-based reasoning (TM-CBR) for supporting green building design[J]. Building and Environment, 2017, 124:388-401.
[23]	TIXIER A J P, HALLOWELL M R, RAJAGOPALAN B, et al. Automated content analysis for construction safety:A natural language processing system to extract precursors and outcomes from unstructured injury reports[J]. Automation in Construction, 2016, 62:45-56.
[24]	SALAMA D M, EL-GOHARY N M. Semantic text classification for supporting automated compliance checking in construction[J/OL]. Journal of Computing in Civil Engineering, 2016, 30(1). https://doi.org/10.1061/(ASCE)CP.1943-5487.0000301.
[25]	汪旭. 建筑质量投诉文本分类与知识问答系统研究[D]. 武汉:华中科技大学, 2018.
[26]	VASWANI A, SHAZEER N, PARMAR N, et al. Attention Is All You Need[M/OL]. arXiv, 2023[2023-08-04]. http://arxiv.org/abs/1706.03762.
[27]	DEVLIN J, CHANG M W, LEE K, et al. BERT:Pre-training of Deep Bidirectional Transformers for Language Understanding[M/OL]. arXiv, 2019[2023-07-20]. http://arxiv.org/abs/1810.04805.
[28]	ZHENG Z, LU X Z, CHEN K Y, et al. Pretrained domain-specific language model for natural language processing tasks in the AEC domain[J/OL]. Computers in Industry, 2022, 142, 103733. https://doi.org/10.1016/j.compind.2022.103733.
[29]	PRIETO S A, MENGISTE E T, GARCÍA DE SOTO B. Investigating the Use of ChatGPT for the Scheduling of Construction Projects[J/OL]. Buildings, 2023, 13(4), 857. https://doi.org/10.3390/buildings13040857.
[30]	UDDIN S M J, ALBERT A, OVID A, et al. Leveraging ChatGPT to aid construction hazard recognition and support safety education and training[J/OL]. Sustainability, 2023, 15(9), 7121. https://doi.org/10.3390/su15097121.
[31]	JI S, PAN S, CAMBRIA E, et al. Survey on knowledge graphs:representation, acquisition, and applications[J]. IEEE Transactions on Neural Networks and Learning Systems, 2022, 33(2):494-514.
[32]	wenda-LLM/wenda:闻达:一个LLM调用平台[EB/OL]//GitHub.[2023-08-09]. https://github.com/wenda-LLM/wenda.
[33]	REIMERS N, GUREVYCH I. Sentence-BERT:Sentence embeddings using siamese BERT-Networks[C/OL]//Proceedings of the 2019 conference on empirical methods in natural language processing. Association for Computational Linguistics, 2019. https://arxiv.org/abs/1908.10084.
[34]	JOHNSON J, DOUZE M, JÉGOU H. Billion-scale similarity search with GPUs[J]. IEEE Transactions on Big Data, 2019, 7(3):535-547.
[35]	TOUVRON H, LAVRIL T, IZACARD G, et al. LLaMA:Open and Efficient Foundation Language Models[M/OL]. arXiv, 2023[2023-07-20]. http://arxiv.org/abs/2302.13971.
[36]	叶列平. 混凝土结构(上册)[M]. 第2版. 北京:中国建筑工业出版社, 2014.