电气测量中的多传感器信息融合技术

Fusion Technique of Multisensor Information in Electronic Measurements

由于电气设备的测量精度有限以及传感器故障或受到噪声、电磁波等多种因素的干扰 ,使得电信号的精确估计存在一定的难度。基于Kalman滤波和多尺度传感器 ,采用局部小波预值处理 ,给出了多传感器系统重构的方法。仿真表明 :该算法不仅利用多传感器提高了估计精度 ,而且对于传感器的突变型未知干扰具有很强的鲁棒性。而这种干扰可能来自于传感器故障或者随机电磁干扰。为在传感器故障或者随机电磁干扰的条件下精确估计系统状态 ,进而分析系统 (包括系统组件故障诊断 ) ,优化运行参数 ,为有效控制提供了技术保证。

Due to the random unknown disturbances or faults in measurement sensors of electronic equipments,encounter the difficulty to estimate the system states accurately for further system analysis,adaptive reconfigurement,and even feedback control.Proposed a new information fusion technique based on Kalman filters and multiscale sensors by adding wavelet preprocessing course.The simulations shown that the new information fusion scheme not only can inprove the accuracy but also have strong robustness to unknown abruptly random disturbance coming from sensor fault or electromagnetic interfere.