基于RBF网络的鱼雷控制系统故障诊断

Fault Diagnosis of Torpedo Control System Based on RBF Neural Network

文中研究了将RBF(径向基)网络应用于鱼雷深度控制系统的故障诊断方法。根据鱼雷深度控制规律,设定五类故障,并应用相关数据对RBF网络进行训练,然后使用满足训练要求的网络对故障数据进行分析,判断属于何类故障。仿真实验结果表明该方法能够有效地解决鱼雷深度控制系统故障诊断问题。

A fault diagnosis method which applies the RBF（radial basis function）NN to the torpedo depth control system is provided. According to the control regulation, set five type faults, and apply the related data to carry on the training to RBF NN, then use the network satisfies the training request to analyze the fault data, and to judge it belongs to which type. The simulation result indicates that this method can efficiently solve the fault diagnosis of the torpedo depth control system.