论文部分内容阅读
研究了神经网络信息融合技术在模拟电路故障诊断中的应用。利用BP神经网络建立信息融合中心,对多传感器数据进行融合处理,减少模拟电路故障诊断的不确定性。对标准电路的输出电压与电源电流特征信息,及两者融合信息的故障诊断性能比较,表明神经网络信息融合方法用于模拟电路故障诊断是有效和可行的。
The application of neural network information fusion technology in analog circuit fault diagnosis is studied. Using BP neural network to establish information fusion center, the fusion of multisensor data is processed to reduce the uncertainty of analog circuit fault diagnosis. The comparison of the output voltage and current characteristics of the standard circuit with the fault diagnosis performance of the fusion information shows that the neural network information fusion method is effective and feasible for analog circuit fault diagnosis.