面向船用薄板焊接的缺陷根因分析框架OA
Root cause analysis framework for welding defects in marine thin plate applications
薄板焊接是船舶建造中的一项核心工艺.然而,该工艺中薄板易产生各类缺陷,为船舶服役埋下安全隐患.针对船用薄板焊接缺陷,提出一套数据驱动型的缺陷全链路诊断框架.首先,提出了用于时空重构的Mamba残差图卷积网络(Mamba Residual Graph Convolutional Network for Spatio-Temporal Reconstruction,MRGSTR),该模型基于重构的方式,对焊接传感器多变量时序数据进行实时无监督异常检测.其次,设计了基于CNN-BiGRU的缺陷检测方法,并探索如何基于贝叶斯网络,融合时序数据与工艺知识图谱进行缺陷根因分析.最后,集成了一个多智能体检索增强生成(Retrieval-Augmented Generation,RAG)的船用薄板焊接缺陷检测运维系统,形成了"传感器实时数据异常检测-缺陷检测-缺陷根因分析-缺陷维修建议"的完整缺陷检测链路,力求为后续工程化提供可演进的路径.
Thin plate welding constitutes a core process in shipbuilding.During this process,thin plates are prone to various defects,posing potential safety hazards during vessel service.This paper proposes a data-driven full-chain diagnostic solution for defects in marine thin plate welding.First,the Mamba Residual Graph Convolutional Network for Spatio-Temporal Reconstruction(MRGSTR)is introduced,which facilitates real-time unsupervised anomaly detection on multivariate time-series(MTS)data from welding sensors based on a reconstruction approach.Subsequently,a CNN-BiGRU-based defect detection method is designed,and Bayesian networks are employed to fuse time-series data with a process knowledge graph for root cause analysis.Finally,a multi-agent Retrieval-Augmented Generation(RAG)system for operation and maintenance of thin plate welding defect detection is integrated,when a comprehensive defect detection chain is established:"Sensor data anomaly detection-Defect identification-Root cause analysis-Repair recommendations",aiming to provide an evolvable pathway for future engineering implementation.
王安民;吉永军;李程炜;夏阳;刘天元;鲍劲松
东华大学机械工程学院 上海 201620上海外高桥造船有限公司 上海 200120东华大学机械工程学院 上海 201620上海外高桥造船有限公司 上海 200120东华大学机械工程学院 上海 201620东华大学机械工程学院 上海 201620
薄板焊接缺陷检测根因分析
thin plate weldingdefect detectionroot cause analysis
《金属加工(热加工)》 2026 (9)
70-79,10
国家自然科学基金面上项目(52475513)国家自然科学基金青年项目(52405536).
评论