摘要
介绍利用径向基神经网络构造了一种在线故障诊断及信号恢复方法,给出了网络的连接结构和学习算法。采用RBF神经网络进行传感器在线故障诊断和信号恢复,其仿真结果表明,该方法具有收敛速度快、信号恢复准确度高、泛化能力强的特点,且可以诊断多种复杂工作系统的传感器在线故障信号,同时进行信号的恢复。实现传感器状态监测、故障诊断、分离和信号恢复。
A new approach of online fault diagnosis and signal restoration formed with RBF neural network is proposed. Also, the network structure and the learning algorithms are given. The RBF neural network is adopted to make online fault diagnosis and signal restoration. The simulation results show that this method has the properties of faster convergence, higher accuracy, better capability of generalization. And online failure of sensor for multiple workingsystems is identified and signal restoration is made. It can meet the needs of CMFD,separation and signal restoration of sensor.
出处
《传感器技术》
CSCD
北大核心
2003年第10期50-53,共4页
Journal of Transducer Technology
关键词
RBF
传感器
在线故障诊断
信号恢复
神经网络
径向基函数
sensor
signal restoration
radial basis function (RBF)
on-line learning
condition monitoring-fault diagnosis (CMFD)
