摘要
针对流程工业中工业安全控制系统硬件模块故障带电诊断困难,且难以及时排查带电情况下的隐性电路故障问题,提出一种实时硬件电路带电故障诊断设计。通过5层小波包分解工业控制系统硬件电路中的突变信号,将提取的信号特征向量结合RBF神经网络训练和分类,能够有效地对工业控制系统硬件电路中的突变故障信号进行识别,达到故障诊断和定位的目的。在故障发生前期能够快速消除隐患,有效保证了工业控制系统运行的安全性和高可靠性。
Aiming at the difficulty in the live diagnosis of hardware module faults of complex industrial control systems in the process industry,and the difficulty of timely troubleshooting the hidden circuit faults under live conditions,a real-time hardware circuit live fault diagnosis design is proposed in this paper,which decomposes the industrial control system hardware through a 5-layer wavelet packet.The mutation signal in the circuit,combining the extracted signal feature vector with RBF neural network training and classification,can effectively identify the mutation fault signal in the hardware circuit of the industrial control system,and achieve the purpose of fault diagnosis and location.It can quickly eliminate hidden dangers in the early stage of failure,effectively guarantee the safety and high reliability of industrial control system operation.
出处
《工业控制计算机》
2023年第7期20-22,25,共4页
Industrial Control Computer
关键词
工业控制系统
小波包分解
特征向量
神经网络
诊断
industrial control system
wavelet packet decomposition
feature vector
neural network
diagnosis