计算机工程与应用2025，Vol.61Issue(11)：227-237,11.DOI:10.3778/j.issn.1002-8331.2403-0143
PCB工艺FMEA领域知识图谱构建与应用
Construction and Application of Knowledge Graph in PCB Process FMEA Field
摘要
Abstract
With the rapid development of electronic products,industrial chain manufacturers have put forward higher requirements for the failure mode and effect analysis(FMEA)of printed circuit boards(PCB),and the traditional FMEA analysis method can no longer meet the needs of complex electronic products.This paper proposes a set of PCB process FMEA knowledge graph construction and application frameworks to realize a novel end-to-end FMEA analysis model.During the graph construction,given the characteristics of a large number of entities in complex sentences,a named entity recognition(NER)model of FLEBERT with PCB and FMEA feature dictionaries added is trained to recognize the failed data,and the experimental comparison proves that the effect is good.The entities after recognition use Sentence-BERT combined with the FLEBERT pre-training model for entity alignment to improve the quality of knowledge.Finally,Neo4j is used for knowledge storage to complete the construction of the knowledge graph.Based on the constructed knowledge graph,this paper builds the FMEA knowledge graph platform,and initially realizes the application of knowledge explora-tion,knowledge Q&A and knowledge recommendation,demonstrating that knowledge graph technology has good applica-tion prospects in the PCB process FMEA analysis field.关键词
知识图谱/印制电路板(PCB)/失效模式和影响分析(FMEA)/命名实体识别(NER)/BERTKey words
knowledge graph/printed circuit boards(PCB)/failure mode and effect analysis(FMEA)/named entity recognition(NER)/BERT分类
信息技术与安全科学引用本文复制引用
叶进,林琦越,唐欣,王秋祥,胡宁..PCB工艺FMEA领域知识图谱构建与应用[J].计算机工程与应用,2025,61(11):227-237,11.基金项目
2021年工业互联网创新发展工程(TC210804F) （TC210804F）
国家自然科学基金区域联合项目(U22A2054). （U22A2054）
