CAJ | 학술논문

针对基于方法的模糊逻辑和传统的神经网络在故障诊断中各自所具有的不足, 提出了一种可行的方法, 即在使用BP算法的神经网络中, 输入矢量由故障症状的隶属度值组成, 而输出矢量由该症状相对于各故障原因的隶属度值组成.将该方法应用于雷达设备的故障诊断, 得到了令人满意的结果.
침대기우방법적모호라집화전통적신경망락재고장진단중각자소구유적불족, 제출료일충가행적방법, 즉재사용BP산법적신경망락중, 수입시량유고장증상적대속도치조성, 이수출시량유해증상상대우각고장원인적대속도치조성.장해방법응용우뢰체설비적고장진단, 득도료령인만의적결과.
In view of respective disadvantages of the fuzzy logic based on methods and the traditional neural networks when they are applied to the fault diagnosis, a practical way is proposed in this paper. That is, in the neural networks using BP algorithm, the input vector consists of membership values of fault symptoms while the output vector consists of membership values of fault symptoms in terms of each fault reason. Applying the method to the fault diagnosis of a radar equipment, satisfactory results have been got.