模糊神经网络在模拟电路故障诊断中的应用  

Application of Fuzzy Neural Network in Fault Diagnosis of Analog Circuits

作　　者：杜文霞[1] 辛涛[1] 孙昊[2] 吕锋[1] 

机构地区：[1]河北师范大学职业技术学院,河北石家庄050031 [2]华北电力大学电气与电子工程学院,河北保定070031

出　　处：《自动化仪表》2009年第1期6-9,共4页Process Automation Instrumentation

基　　金：国家自然科学基金(编号:60774070);河北省自然科学基金(编号:F2008000197)

摘　　要：将模糊理论和神经网络结合起来用于模拟电路的故障诊断是一种有效的方法。首先测得模拟电路在故障状态和正常状态时各测试节点的电压值,并对其电压偏差值模糊化,从而求得不同故障状态下各测试节点的电压偏差值所对应的隶属度;其次,以各测试节点的隶属度和所对应的故障类型作为训练神经网络的样本,来建立从故障特征到故障类型的非线性函数;最后,将该方法用于视频放大电路的故障诊断。仿真结果验证了该方法的有效性。Combining fuzzy theory and neural network is an effective way to be applied in fault diagnosis of analog circuits. Firstly, the values of voltage on various test points under normal states and fault states are measured ; and the deviation of voltage is fuzzed for deriving the membership degree of the deviation under different fault states. Then,with the membership degree of each test points and their relevant fault categories as the sample for training neural network to establish non-linear function from fault feature to fault categories. Finally, the method is used for fault diagnosis of video amplifying circuit. The result of simulation verified the effectiveness of this method.

关 键 词：模糊理论 神经网络 隶属函数 故障特征提取 故障诊断 模拟电路 

分 类 号：U226.81[交通运输工程—道路与铁道工程]