基于改进RBF神经网络的某型机载雷达电路故障诊断

Fault Diagnosis for a Certain Type of Airborne Radar Circuit Based on Improved RBF Neural Networ

RBF神经网络在故障诊断中的应用是一个十分热门的研究课题。但RBF神经网络用于故障诊断需要大量的隐含层节点,隐含层节点越多其训练时间越长。针对此问题,提出了一种改进的RBF神经网络故障诊断方法。通过仿真实例,可以看出隐含层的节点大大减少,训练时间缩短,该方法用于雷达故障诊断非常有效。

Study on RBF neural network applied in diagnosing fault is one of important problems.However,there are many nerve cells of hidden layer when RBF neural network is applied in diagnosing fault.Training time is longer when nerve cells of hidden layer are more.A kind of diagnosing fault's method based on improved RBF neural network is put forward to solve the problem.The nerve cells of hidden layer is less and Training time is shorter.The method is effective when it is applied in diagnosing the faults of radar's circuits.