摘要
控制棒驱动机构(CRDM)作为核电厂的关键执行部件,通过电磁-机械耦合控制多组线圈电流信号,以实现控制棒精准动作,其健康状态关乎反应堆安全运行。然而,现有CRDM异常检测方法主要聚焦于单组线圈建模,未充分考虑多线圈间的变化规律和动态模式。为此,提出一种多线圈联合学习(MCJL)模型,旨在精准检测控制棒动作过程中的潜伏性异常。首先,定义线圈节点和全连接边,并引入衰减邻接矩阵,以构建多线圈联动图结构,进而表征CRDM内部提升线圈、移动线圈和保持线圈间的动态耦合关系。其次,采用移动图卷积网络,以高效捕获多线圈间的局部时序依赖关系,并协同重构3组线圈的电流信号。最后,通过计算重构信号和真实信号间的残差,利用多尺度动态检测策略实现CRDM逐个动作周期的异常检测。对某地区压水堆机组的历史监测数据与模拟异常样本进行算例分析,表明该方法可有效捕获多线圈间的动态耦合关系并提取驱动机构动作的时序特征,实现电流信号的精准重构。相较于多种异常检测方法,所提MCJL模型在信号重构和异常检测方面具有更优越的性能,其动态阈值策略可灵活调整决策边界,具有较强的容错能力。
The control rod drive mechanisms(CRDMs),critical actuators in nuclear power plants,manipulate multiple coil currents through electromagnetic-mechanical coupling to enable precise rod motion.Their operational health is essential for reactor safety.However,existing anomaly detection methods focus on single-coil modeling,overlooking inter-coil dynamics and evolving patterns.Thus,we propose a multi-coil joint learning(MCJL)model for accurately detecting latent anomalies during control rod movement.First,coil nodes and fully connected edges are defined,with a decay adjacency matrix introduced to construct a multi-coil interaction graph,which captures dynamic coupling among lift,moving,and solid coils.Then,the moving graph convolutional network efficiently extracts local temporal dependencies among coils while jointly reconstructing current signals from all three coils.Finally,residuals between reconstructed and real signals are computed,enabling per-cycle anomaly detection via a multi-scale dynamic strategy.Tests on historical data and simulated anomalies from a pressurized water reactor show the method effectively captures dynamic coil coupling and extracts temporal features,achieving precise signal reconstruction.Compared to existing methods,MCJL excels in reconstruction and detection,with its dynamic threshold strategy offering flexible decision boundaries and strong fault tolerance.Control rod drive mechanisms(CRDMs),as critical actuators in nuclear power plants,regulate control rod motion by manipulating multiple coil currents through electromagnetic-mechanical coupling.Their operational integrity is vital for ensuring reactor safety.However,existing anomaly detection approaches predominantly model individual coils,neglecting the dynamic interactions among coils and the evolving operational patterns.To address this gap,we propose a multi-coil joint learning(MCJL)model for accurately detecting latent anomalies during control rod operation.In this approach,coil nodes and fully connected edges are defined,and a decay adjacency matrix is introduced to construct a multi-coil interaction graph that captures the dynamic coupling among the lift,moving,and solid coils.A moving graph convolutional network is then employed to efficiently extract local temporal dependencies across coils while jointly reconstructing the current signals of all three coils.The residuals between the reconstructed and actual signals are subsequently calculated,enabling per-cycle anomaly detection using a multi-scale dynamic strategy.Experimental validation using historical data and simulated anomalies from a pressurized water reactor demonstrates that the proposed method effectively captures dynamic inter-coil coupling and temporal patterns,achieving high-precision signal reconstruction.Compared with existing methods,MCJL exhibits superior performance in both reconstruction and anomaly detection.Furthermore,its dynamic thresholding strategy provides flexible decision boundaries and strong fault tolerance.
作者
林蔚青
缪希仁
江灏
叶铭新
陈静
Lin Weiqing;Miao Xiren;Jiang Hao;Ye Mingxin;Chen Jing(College of Electrical Engineering and Automation,Fuzhou University,Fuzhou 350108,China)
出处
《仪器仪表学报》
北大核心
2025年第6期290-303,共14页
Chinese Journal of Scientific Instrument
基金
福建省高校产学合作项目(2023H6006)资助。
关键词
核电厂
控制棒驱动机构
异常检测
图卷积网络
动态阈值
nuclear power plant
control rod drive mechanism
anomaly detection
graph convolution network
dynamic threshold
