自动舵电路板神经网络故障诊断方法

Research on circuit boards fault diagnosis of auto rudder based on neural network

为了实现对自动操舵仪电路故障的自动检测和诊断、达到降低维修成本和提高维修效率的目的,采用神经网络的方法,通过在电路板的输入端输入不同的测试序列,仅在电路板的输出端进行数据采集和特征提取,然后利用神经网络进行故障诊断,从而达到自动判断电路板好坏和定位常见故障元件的目的,避免使用探笔或夹具时因接错位置而不能定位故障的问题。采用该方法建立的故障诊断系统操作简单,使用方便,具有智能化、可视化的特点。

To test and diagnose circuit board of a rudder automatically,so that the maintain cost would be reduced and the efficiency would be improved,a neural network fault diagnosis method was developed.We input a serial of different signals to the entrance of the circuit board,acquire data from the terminal of the circuit board,extract characteristic of the data as the inputs of a neural network to ascertain the fault,and diagnosis the circuit board by neural network.Then we can find out whether the board is good or not and indicate the bad component of the board.It can avoid the problem that caused by wrong insertion of probes and clamps,and make the diagnosis easier for users,especially for untrained users.We can develop an operable,visible and intellectualized diagnosis system by this method.