通过模态滤波实现阵列式传感器系统的故障诊断

摘要

By using the modal filtering method,a modified sensor fault detection approach based on structural dynamic characteristics was presented.An array of accelerometers attached on the surface of a beam was taken as sensors. The differences between the reconstructed signals with modal filtering from the output signals of the beam system and the real measured signals were obtained,then the curvature errors were taken as sensitive indexes to detect and identify the simulated fault sensor of the sensor system.The numerical simulations and experimental results showed that the proposed method can be used to detect the fault sensor system efficiently;furthermore,this method is independent upon the location of an excitation and it can be used for real-time monitoring systems.%基于结构振动响应特性利用改进的模态滤波方法对阵列式传感器系统进行故障诊断。在梁结构表面均匀布置一组加速度计，利用模态振型对该系统的输出信号进行重构，将重构信号与实际信号之间的曲率误差作为敏感参数，对系统中的模拟故障传感器进行检测与识别，并加以实验验证。数值计算和实验结果表明：改进的模态滤波方法不仅可以直接有效地对传感器系统进行实时故障监测，而且该方法与外界激励力位置无关，具有良好的工程应用前景。