Research on Machining Process Recommendation Method of Structural Parts Based on Large Language Models

doi:10.3901/JME.2025.17.393

Abstract

Abstract: In the process of numerical control programming for complex structural components, the difficulty in reusing machining process knowledge arises due to the heterogeneity of knowledge sources and the complexity of interconnections between knowledge. A knowledge recommendation method for structural parts machining process based on a large language model is proposed. By selecting and fine-tuning the large language model, a vertical domain model of machining process knowledge recommendation for structural parts is established. The evaluation results indicate that the model can recommend corresponding machining processes based on specific part features. To solve the problem of the model not being able to obtain the latest professional knowledge and the low accuracy of machining process recommendations, the LangChain application framework combined with a knowledge base is used to enhance the knowledge retrieval of the domain model and construct a process knowledge question answering system. Through corresponding indicator evaluation, the F1 value of the question answering system improves by 0.026 on the basis of the original domain model, and the accuracy of machining process recommendations is above 90%. In the process decision-making application of CNC programming for aviation structural components, this method recommends corresponding process knowledge based on part features. Compared with the automatic CNC programming system that does not use the method in this article, the efficiency of generating CNC codes for frame type structural components improves to a certain extent, which is of great significance for improving the decision-making efficiency of CNC programmers.

Key words: large language models, machining process, knowledge recommendation, rapid NC programming

CLC Number:

TH16

ZHENG Xiaohu, CHEN Hongbo, HE Fangzhou. Research on Machining Process Recommendation Method of Structural Parts Based on Large Language Models[J]. Journal of Mechanical Engineering, 2025, 61(17): 393-404.

References

[1] 尹佳, 唐宇阳, 张俊, 等. 基于复合加工特征的航空结构件频响快速预测[J]. 机械工程学报, 2023, 59(3): 200-207. YIN Jia, TANG Yuyang, ZHANG Jun, et al. Rapid frequency response function prediction of aeronautical structural parts based on composite machining features[J]. Journal of Mechanical Engineering, 2023, 59(3): 200-207.
[2] 常智勇, 陶礼尊, 李佳佳, 等. 基于加工意图的机加工艺知识重用方法研究[J]. 机械工程学报, 2018, 54(3): 160-168. CHANG Zhiyong, TAO Lizun, LI Jiajia, et al. The measure and search method of process knowledge element based on machining intent[J]. Journal of Mechanical Engineering, 2018, 54(3): 160-168.
[3] ZHOU D, DAI X. A method for discovering typical process sequence using granular computing and similarity algorithm based on part features[J]. The International Journal of Advanced Manufacturing Technology, 2015, 78(9): 1781-1793.
[4] 罗滨鸿, 周虎, 张祺薇, 等. 基于Apriori算法的网线编织工艺缺陷数据挖掘方法[J]. 制造业自动化, 2022, 44(5): 75-77+102. LUO Binhong, ZHOU Hu, ZHANG Qiwei, et al. A data mining method for defect data of net rope weaving process based on apriori algorithm[J]. Manufacturing Automation, 2022, 44(5): 75-77+102.
[5] 阳树梅, 王华昌, 李建军. 基于经验的数控工艺知识挖掘算法研究[J]. 模具工业, 2023, 49(8): 1-10. YANG Shumei, WANG Huachang, LI Jianjun. Research on knowledge mining algorithm of NC process based on experience[J]. Die & Mould Industry, 2023, 49(8): 1-10.
[6] ZHOU B, BAO J S, LI J, et al. A novel knowledge graph-based optimization approach for resource allocation in discrete manufacturing workshops[J]. Robotics and Computer-Integrated Manufacturing, 2021, 71(3): 102160-102173.
[7] ZHENG P, XIA L Q, LI C X, et al. Towards Self-X cognitive manufacturing network: an industrial knowledge graph-based multi-agent reinforcement learning approach[J]. Journal of Manufacturing Systems, 2021, 61(1): 16-26.
[8] ZHOU B, LI X, LIU T, et al. CausalKGPT: industrial structure causal knowledge-enhanced large language model for cause analysis of quality problems in aerospace product manufacturing[J]. Advanced Engineering Informatics, 2024, 59: 102333.
[9] 夏润泽, 李丕绩. ChatGPT大模型技术发展与应用[J].数据采集与处理, 2023, 38(5): 1017-1034. XIA Runze, LI Piji. Large language model ChatGPT: evolution and application[J]. Journal of Data Acquisition and Processing, 2023, 38(5): 1017-1034.
[10] DING N, QIN Y J, YANG G, et al. Parameter-efficient fine-tuning of large-scale pre-trained language models[J]. Nature Machine Intelligence, 2023, 5(3): 220-235.
[11] KOJIMA T, GU S S, REID M, et al. Large language models are zero-shot reasoners[J]. Advances in Neural Information Processing Systems, 2022, 35: 22199-22213.
[12] LIU Xiao, JI Kaixuan, FU Yicheng, et al. P-Tuning v2: prompt tuning can be comparable to fine-tuning universally across scales and tasks[M/OL]. arXiv, 2021[2023-12-6]. https: //arxiv.org/abs/2110.07602.
[13] HU E J, SHEN Y, WALLIS P, et al. Lora: Low-rank adaptation of large language models[M/OL]. arXiv, 2021[2023-12-1]. https: //arxiv.org/abs/2106.09685.
[14] DETTMERS T, PAGNONI A, HOLTZMAN A, et al. Qlora: efficient finetuning of quantized llms[M/OL]. arXiv, 2023[2023-12-8]. https: //arxiv.org/abs/2305.14314.
[15] 张鹤译, 王鑫, 韩立帆, 等. 大语言模型融合知识图谱的问答系统研究[J]. 计算机科学与探索, 2023, 17(10): 2377-2388. ZHANG Heyi, WANG Xin, HAN Lifan, et al. Research on question answering system on joint of knowledge graph and large language models[J]. Journal of Frontiers of Computer Science and Technology, 2023, 17(10): 2377-2388.
[16] 赵鑫, 窦志成, 文继荣. 大语言模型时代下的信息检索研究发展趋势[J]. 中国科学基金, 2023, 37(5): 786-792. ZHAO Xin, DOU Zhicheng, WEN Jirong. The development of information retrieval in the era of large language model[J]. Bulletin of National Natural Science Foundation of China, 2023, 37(5): 786-792.
[17] SINGHAL K, AZIZI S, TU T, et al. Large language models encode clinical knowledge[J]. Nature, 2023, 620(7972): 172-180.
[18] HUAGN A H, WANG H, YANG Y. FinBERT: A large language model for extracting information from financial text[J]. Contemporary Accounting Research, 2023, 40(2): 806-841.
[19] 张俊, 徐箭, 许沛东, 等. 人工智能大模型在电力系统运行控制中的应用综述及展望[J]. 武汉大学学报(工学版), 2023, 56(11): 1368-1379. ZHANG Jun, XU Jian, XU Peidong, et al. Overview and prospect of application of artificial intelligence large model in power system operation control[J]. Engineering Journal of Wuhan University, 2023, 56(11): 1368-1379.
[20] TOUVRON H, LAVRIL T, IZACARD G, et al. Llama: open and efficient foundation language models[M/OL]. arXiv, 2023[2023-12-20]. https: //arxiv.org/abs/2302.13971.
[21] RADFORD A, WU J, CHILD R, et al. Language models are unsupervised multitask learners[J]. OpenAI blog, 2019, 1(8): 9.
[22] TAORI R, GULRAJANI I, ZHANG Tianyi, et al. Stanford alpaca: An instruction-following llama model[EB/OL]. GitHub Repository, 2023[2023-12-12]. https: //github.com/tatsu-lab/stanford_alpaca.
[23] DU Zhengxiao, QIAN Yujie, LIU Xiao, et al. General language model pretraining with autoregressive blank infilling[C]//Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics. Dublin: ACL, 2022: 320-335.
[24] GUU K, LEE K, TUNG Z, et al. Retrieval augmented language model pre-training[C]//Proceedings of the 37th International Conference on Machine Learning. PMLR, 2020: 3929-3938.
[25] PAPINENI K, ROUKO S, WARD T, et al. Bleu: a method for automatic evaluation of machine translation[C]//Proceedings of the 40 th annual meeting of the Association for Computational Linguistics. Philadelphia, Pennsylvania, USA: ACL, 2002: 311-318.
[26] LIN C Y. Rouge: A package for automatic evaluation of summaries[C]//Text summarization branches out. Barcelona, Spain: ACL, 2004: 74-81.
[27] KORBAK T, ELSAHAR H, KRUSZEWSKI G, et al. Controlling conditional language models without catastrophic forgetting[C]//Proceedings of the 39th International Conference on Machine Learning. Baltimore, Maryland, USA: PMLR, 2022, 162: 11499-11528.
[28] ZHANG Tianyi, KISHORE V, WU F, et al. Bertscore: evaluating text generation with bert[M/OL]. arXiv, 2019[2023-12-26]. https: //arxiv.org/abs/1904.09675.