摘要
为了提高模拟电路故障诊断的准确率,要尽可能的采集有效故障样本信息作为BP神经网络的输入。提出利用输出电压和电源电流信息融合的方法进行模拟电路故障诊断。收集输出电压和电源电流的故障样本集,然后作为BP神经网络的输入对网络进行训练和判断。利用不同故障对输出电压和电源电流的影响不同,能减少故障特征的重叠,提高模拟电路的故障诊断正确率。仿真结果表明利用输出电压和电源电流信息融合的方法比单纯利用输出电压或电源电流进行诊断准确率和速度都有明显提高。
In order to improve the accuracy of analog circuit fult diagnosis ,more effective fault samples information need to be collected as far as possible as input of BP neural network .A method of using output voltage and electric current information fusion for analog circuit fault diagnosis is put forward .The output voltage and electric current of the fault sample set is collected and then as a BP neural network input of network training and judgment .Using different fault impact on out‐put voltage and electric current is different ,the fault characteristics of overlap is reduced and the accuracy of analog circuit fault diagnosis is improved .The simulation results show that the method using output voltage and electric current informa‐tion fusion accuracy and speed are improved obviously than just using the output voltage or power supply current diagnosis .
出处
《舰船电子工程》
2015年第6期118-121,共4页
Ship Electronic Engineering
关键词
信息融合
BP
神经网络
模拟电路
故障诊断
information fusion
BP neural network
analog circuits
fault diagnosis