基于RBF神经网络的传感器故障诊断研究

Research on sensor fault diagnosis based on RBF neural network

提出一种基于RBF神经网络的多传感器故障在线诊断方法..同时对该方法在摆式列车 加速度传感器故障诊断中的应用进行了仿真研究。仿真结果表明此方法是有效的,不仅消除了零偏对 故障诊断的影响,而且提高了故障检测率及可靠性。

An in-line fault diagnosis method for multi-sensors based on RBF neural network is brought forward. Simulations are made for its application to the acceleration sensor of the tilting train, which indicates the method is active. The influences to the fault diagnosis by zero deviation are eliminated and the fault detection rate is improved as well as the reliability.