知识图谱协同大模型赋能制造业质量管理的智能化转型

Decoding the Intelligent Transformation of Quality Management in Manufacturing with Collaborative Knowledge Graphs and Large Language Models

在制造业智能化奏响攻坚号角的背景下,质量管理的智能化转型已是必然趋势。然而,当前制造业在质量管理智能化转型过程中遇到了诸多问题,比如知识碎片化、知识体系不完整和经验个人化等。用于解决这些问题的知识图谱和大模型等技术同样面临知识抽取、可信问答等挑战。文章提出知识图谱协同大模型的方法,用于很好地解决质量管理智能化过程中的知识图谱构建、知识的可信问答以及针对质量管理普遍存在的智能化归因分析等所面临的挑战。接着,文章给出三个具有普遍示范性的案例,包括汽车维修中的问答式引导诊断、精密电子设备故障的多维分析和电信网络的智能运维与故障诊断。最后,对大模型时代制造业的智能化转型进行展望。

Amidst the all-out push for intelligent manufacturing,the intelligent transformation of quality management has become an inevitable trend.However,during the current process of intelligent transformation of quality management in the manufacturing industry,numerous problems have been encountered,such as fragmented knowledge,incomplete knowledge system and personalized experience.Technologies like knowledge graphs and large language models,which are used to solve these problems,also face challenges like knowledge extraction and trustworthy question answering.This paper proposes a method of collaborative knowledge graphs and large language models to effectively address the challenges faced in knowledge graph construction,trustworthy question answering and intelligent attribution analysis for quality management during the intelligent transformation process.Furthermore,the paper presents three universally demonstrative cases,including question-and-answer-guided diagnosis for automobile repairs,multi-dimensional analysis of precision electronic equipment failures and intelligent operation and maintenance and fault diagnosis for telecommunication networks.Finally,the paper provides an outlook on the intelligent transformation of the manufacturing industry in the era of large language models.