基于神经网络传感器组改障诊断方法研究

对于系统状态检测与故障诊断,传感器自身故障的及早发现很重要.本文提出了一种基于神经网络和冗余率计算的传感器组故障监测与诊断的新方法.该方法先用RBF神经网络对传感器组中的各个输出进行预测,若预测值与输出值发生较大的偏差,可能是传感器故障或设备故障,运用传感器之间的冗余率,进一步判断传感器是否发生故障,进而采用对应的诊断策略.

The Study Of The Fault Diagnosis For Inspection System Fixing With The Redundant Transducer Group Based On RBF Neural Network

It is very important to find the fault of the transducer itself for the system inspection and monitoring. A transducer fault diagnosis method based on the RBF neural network and the redundancy calculation is put forward. Firstly, it uses the RBF neural network to predict the output of the transducer. If the error between the prediction and the actual output goes beyond the limit, it is possibility of the transducer or the mechanism being malfunction. Then with the help of calculation of the redundancy between the transducers, the fault is judged more exactly, the diagnosis strategy is come into being in succession.