基于RBF神经网络和定性趋势分析的传感器故障诊断

Research on Sensor Fault Diagnosis Based on RBF Neural Network Model and Qualitative Trend Analysis

针对传感器偏置故障及漂移故障,提出了一种基于RBF神经网络和定性趋势分析的传感器故障诊断方法。该方法充分利用控制系统闭环回路测控信息,建立RBF神经网络预测器,通过将RBF神经网络的预测输出值与传感器实际输出相比较获取残差序列,根据残差首先判断传感器是否发生故障,然后用定性趋势分析方法获得传感器偏置故障和漂移故障的辨识策略,实现传感器故障在线识别。应用结果表明该方法不仅可以提高传感器预测精度,而且可以快速准确地检测和辨识传感器故障类型及故障发生时间。

Aiming at the issues of sensor bias and drift faults,a novel approach of sensor fault diagnosis based on RBF neural network model and qualitative trend analysis is proposed.With the method,the closed-loop monitoring information in control system is adopted to establish RBF neural network forecaster.By comparing the forecasting outputs of the RBF neural network model and the actual values of sensor,the identification strategy based on the sequence of residuals for sensors bias fault and drift fault is acquired and on-line sensors fault diagnosis is carried out.The application results indicate that the approach can not only improve the forecasting accuracy of sensor but also check and identify the type and occurrence time of the fault quickly and accurately.